CN119884909B

CN119884909B - Melting monitoring system and method for solid coloring agent

Info

Publication number: CN119884909B
Application number: CN202510308589.XA
Authority: CN
Inventors: 王炳林; 毛林荣; 汪春水; 王水炳
Original assignee: Zhejiang Weifeng New Material Co ltd
Current assignee: Zhejiang Weifeng New Material Co ltd
Priority date: 2025-03-17
Filing date: 2025-03-17
Publication date: 2025-06-10
Anticipated expiration: 2045-03-17
Also published as: CN119884909A

Abstract

The embodiment of the present application provides a melting monitoring system and method for solid colorants. It includes setting a light source and a light sensor to generate an initial optical path, recording and processing light intensity values to form a time-light intensity curve, using a support vector machine model to identify the melting starting point, and calculating the melting rate through a Kalman filter to generate a quality assessment report. When the melting rate reaches a preset threshold, the PID controller triggers a cooling program, and combines deep learning technology to analyze the uniformity and purity of the colorant, and finally generates a melting monitoring report. The technical solution provided by the embodiment of the present application improves the product quality and production efficiency of solid colorants.

Description

Melting monitoring system and method for solid coloring agent

Technical Field

The embodiment of the application relates to the technical field of automatic control, in particular to a melting monitoring system and method for a solid coloring agent.

Background

In many industries, such as textiles, plastics, paints, etc., the melting process of solid colorants is a critical step in ensuring the quality of the final product. Accurate monitoring of the melt state of solid colorants is critical to control the production process, optimize process parameters, and ensure product consistency.

Currently, there are several techniques that attempt to solve the problem of solid stain melt monitoring. For example, an optical sensor may be used to record the change in intensity of light transmitted through the stain sample, form a time-light intensity curve, or a temperature sensor may be used to monitor the change in temperature within the heating vessel.

The existing schemes suffer from several significant drawbacks, firstly, they mostly rely on simple threshold decisions or linear regression models to identify melting start points and rates, which appear to be frustrating in the face of complex nonlinear melting processes, prone to false positives. Second, although the conventional denoising method (such as mean filtering) can reduce noise interference, key feature information may be blurred at the same time, which affects the accuracy of subsequent analysis. Finally, the existing system has little intelligent regulation function, and can not automatically regulate production process parameters (such as cooling program) according to real-time monitoring data, so that the application potential of the system in an automatic production line is limited.

Disclosure of Invention

The embodiment of the application provides a melting monitoring system and method for a solid coloring agent, which are used for solving the problems of defective product quality and low production efficiency of the solid coloring agent in the prior art.

In a first aspect, an embodiment of the present application provides a method for monitoring the melting of a solid colorant, including:

Placing a solid dye to be monitored in a transparent heating container, arranging at least one light source on one side of the transparent heating container, arranging at least one light sensor on the opposite side, selecting a wavelength combination according to the absorption spectrum of the solid dye, and utilizing a multi-wavelength light source to enable a light path to penetrate through the maximum section of the solid dye to generate an initial optical path;

Recording the light intensity value received by the light sensor and transmitted through the solid dye according to a preset time interval to form an initial time-light intensity curve, and denoising the data of the initial optical path by a wavelet transformation algorithm to obtain a time-light intensity curve;

Based on the change trend of the time-light intensity curve, observing by using a preset support vector machine model to obtain a turning point of which the slope of the light intensity curve changes, and indicating the moment when the solid coloring agent starts to melt based on the turning point to generate a melting start signal;

Calculating a melting rate by using a Kalman filter based on the data of a time-light intensity curve of the melting start signal in a fixed time, and predicting the stability of the solid coloring agent under different environmental conditions by combining a thermodynamic analysis algorithm to generate a quality evaluation report;

When the melting rate is detected to reach a preset threshold, a PID controller is utilized to automatically trigger a cooling program, the uniformity and purity of the solid coloring agent are predicted and analyzed by combining a deep learning image analysis technology, key parameters are obtained, key parameters and quality evaluation results of the solid coloring agent in the melting process are integrated, and a melting monitoring report is generated.

Optionally, based on the trend of the time-light intensity curve, observing with a preset support vector machine model to obtain a turning point at which the slope of the light intensity curve changes, and based on the turning point, indicating the time when the solid colorant starts to melt to generate a melting start signal, where the method includes:

Smoothing the time-light intensity curve by using a five-point three-time smoothing algorithm to obtain a smooth curve, calculating the slope of the smooth curve according to a center difference method and a least square fitting technology, and obtaining slope values corresponding to all time points;

Performing dimension reduction processing on the slope value by using principal component analysis to obtain a dimension reduction slope value, inputting the dimension reduction slope value into a pre-trained support vector machine model, classifying and identifying the dimension reduction slope value, and obtaining an initial turning point at which the slope of a light intensity curve changes;

based on the initial turning point, matching a time-light intensity curve within a specified time period by using a dynamic time warping algorithm, monitoring model output by using an anomaly detection algorithm, and predicting and positioning the turning point of the change of the slope of the light intensity curve;

based on the turning point, a Kalman filter is utilized to optimize the estimation of the melting start time, and a melting start signal is generated.

Optionally, performing a dimension reduction process on the slope value by using principal component analysis to obtain a dimension-reduced slope value, including:

Optimizing the slope value by using a robust regression algorithm, organizing to form a high-dimensional slope value matrix, and standardizing the characteristics of the high-dimensional slope value matrix according to a Z-Score standardization method to generate a standardized slope value matrix, wherein the characteristics comprise zero mean and unit variance;

Calculating a covariance matrix based on a Markov distance according to the standardized slope value matrix, and accelerating the decomposition process of the covariance matrix by adopting a random singular value decomposition algorithm to obtain a characteristic value and a characteristic vector;

Based on the characteristic value and the characteristic vector, evaluating the performance of the support vector machine model under different principal component numbers by using a cross verification technology, selecting N principal components which can retain the maximized information and verify the performance of the support vector machine model, and screening principal components which are helpful to classification tasks by adopting minimum absolute shrinkage and a selection operator regression when determining the N value, so as to generate a principal component set;

And performing preliminary dimension reduction on the standardized slope value matrix by using a t distribution random neighborhood embedding algorithm to obtain a preliminary dimension reduction slope value, and projecting the preliminary dimension reduction slope value onto a new coordinate system defined by the main component set to generate a dimension reduction slope value.

Optionally, when the melting rate is detected to reach a preset threshold, automatically triggering a cooling program by using a PID controller, and predicting and analyzing the uniformity and purity of the solid dye by combining a deep learning image analysis technology to obtain key parameters, and integrating the key parameters and quality evaluation results of the solid dye in the melting process to generate a melting monitoring report, including:

When the melting rate reaches a preset threshold, optimizing parameter setting of a PID controller by using fuzzy logic, and automatically triggering a cooling program according to a preset temperature control strategy;

in the cooling process, acquiring an initial image of the solid dye by using a high-resolution camera, and processing the initial image based on a multi-scale image fusion technology to obtain an optimized image;

preprocessing the optimized image according to a super-pixel segmentation algorithm to obtain a processed image, and analyzing the characteristics of color distribution, particle size, distribution and the like extracted from the processed image by utilizing an integrated learning method to obtain key parameters including uniformity index, purity grade and impurity content of the solid coloring agent;

according to key parameters, collecting external data in the melting process, and predicting the melting rate and environmental condition change in a fixed time in the future based on time series analysis to obtain an analysis result, wherein the external data comprises the melting rate, the environmental temperature and the humidity;

Based on the analysis results, a natural language processing technique is applied to automatically generate a melt monitoring report.

Optionally, preprocessing the optimized image according to a super-pixel segmentation algorithm to obtain a processed image, including:

Performing weighted average fusion processing on the initial image of the solid coloring agent acquired by the high-resolution camera by utilizing a multi-view data set and a structural similarity index to generate an optimized image;

Initializing a super-pixel grid according to the optimized image, and carrying out iterative optimization processing on the boundary of the super-pixel grid based on a modified version of simple linear iterative clustering algorithm in combination with a spectral clustering technology to obtain a primary segmentation result;

Based on the preliminary segmentation result, an active contour model and a graph cut algorithm are applied to refine and smooth the boundary of the super-pixel grid, and meanwhile, shape regularity constraint such as ellipse fitting is introduced to generate an intermediate segmentation image;

And removing isolated areas by adopting a connected component marking algorithm according to the intermediate segmentation image, and performing morphological operation by applying bilateral filtering and smoothing operation to generate a processed image, wherein the morphological operation comprises an open operation and a closed operation.

Optionally, calculating the melting rate using a kalman filter based on data of a time-light intensity curve of the melting start signal within a fixed time, includes:

Calculating a fixed time when the solid coloring agent starts to melt by utilizing the melting start signal and combining an adaptive threshold algorithm, adopting a dynamic window selection method based on the fixed time, and automatically adjusting the time period length according to the change trend of the melting rate so as to extract initial time-light intensity curve data from the fixed time;

Removing noise and abnormal values of the initial time-light intensity curve data by using a median filtering method, simultaneously introducing a local weighted regression scattered point smoothing technology to retain characteristic information in the initial time-light intensity curve data, and further automatically detecting and removing abnormal values by using a robust statistical method to obtain smoothed time-light intensity curve data;

Based on the smooth time-light intensity curve data, reducing noise interference by using a five-point cubic smoothing algorithm, and calculating a slope value corresponding to each time point by using a least square fitting technology and combining an adaptive weight adjustment strategy, wherein the slope value is expressed as a primary melting rate;

and optimizing the primary melting rate by using an unscented Kalman filter to obtain the melting rate.

Optionally, removing noise and outliers of the initial time-light intensity curve data by using a median filtering method, simultaneously introducing a local weighted regression scattered point smoothing technology to retain characteristic information in the initial time-light intensity curve data, and further automatically detecting and removing outliers by using a robust statistical method to obtain smoothed time-light intensity curve data, including:

performing preliminary denoising on the initial time-light intensity curve data by adopting a median filtering method, and replacing each data point with a median in the neighborhood to reduce noise influence, so as to obtain preliminary denoised data;

Based on the preliminary denoising data, introducing a local weighted regression scattered point smoothing technology, smoothing a curve by carrying out weighted least square fitting on local data points, retaining characteristic information and local variation trend, and generating smoothed curve data;

According to the smoothed curve data, automatically detecting and removing abnormal values by using a robust statistical method to obtain a detection result;

and integrating the detection result, the data after preliminary denoising and the smoothed curve data into smoothed time-light intensity curve data.

In a second aspect, embodiments of the present application provide a melt monitoring system for a solid colorant, comprising:

the generating module is used for placing the solid coloring agent to be monitored in a transparent heating container, arranging at least one light source on one side of the transparent heating container, arranging at least one light sensor on the other opposite side of the transparent heating container, selecting a wavelength combination according to the absorption spectrum of the solid coloring agent, enabling a light path to penetrate through the maximum section of the solid coloring agent by utilizing a multi-wavelength light source, and generating an initial optical path;

The denoising module is used for recording the light intensity value received by the light sensor and transmitted through the solid dye according to a preset time interval to form an initial time-light intensity curve, and performing denoising processing on the data of the initial optical path by a wavelet transformation algorithm to obtain a time-light intensity curve;

the indicating module is used for observing by using a preset support vector machine model based on the change trend of the time-light intensity curve to obtain a turning point at which the slope of the light intensity curve changes, and indicating the moment when the solid coloring agent starts to melt based on the turning point to generate a melting starting signal;

The evaluation module is used for calculating the melting rate by using a Kalman filter based on the data of the time-light intensity curve of the melting start signal in the fixed time, predicting the stability of the solid coloring agent under different environmental conditions by combining a thermodynamic analysis algorithm, and generating a quality evaluation report;

And the monitoring module is used for automatically triggering a cooling program by utilizing a PID controller when the melting rate reaches a preset threshold value, predicting and analyzing the uniformity and purity of the solid coloring agent by combining a deep learning image analysis technology to obtain key parameters, and integrating the key parameters and quality evaluation results of the solid coloring agent in the melting process to generate a melting monitoring report.

In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component, where the storage component stores one or more computer instructions, and the one or more computer instructions are configured to be invoked and executed by the processing component to implement a method for monitoring melting of a solid colorant according to the first aspect.

In a fourth aspect, embodiments of the present application provide a computer storage medium storing a computer program which, when executed by a computer, implements a melt monitoring method for a solid colorant as described in the first aspect.

In the embodiment of the application, a solid coloring agent to be monitored is placed in a transparent heating container, at least one light source is arranged on one side of the transparent heating container, at least one light sensor is arranged on the other opposite side of the transparent heating container, wavelength combination is selected according to the absorption spectrum of the solid coloring agent, and a multi-wavelength light source is utilized to enable a light path to penetrate through the maximum section of the solid coloring agent so as to generate an initial optical path; recording the light intensity value of the solid dye received by the light sensor according to a preset time interval, forming an initial time-light intensity curve, carrying out denoising treatment on the data of the initial optical path by a wavelet transformation algorithm to obtain a time-light intensity curve, observing by a preset support vector machine model based on the change trend of the time-light intensity curve to obtain a turning point with the change of the slope of the light intensity curve, indicating the moment when the solid dye starts to melt based on the turning point to generate a melting start signal, calculating the melting rate by using a Kalman filter based on the data of the time-light intensity curve of the melting start signal in a fixed moment, predicting the stability of the solid dye under different environmental conditions by combining a thermodynamic analysis algorithm to generate a quality evaluation report, automatically triggering a cooling program by using a PID controller when the melting rate reaches a preset threshold value, predicting and analyzing the uniformity and purity of the solid dye by combining the image analysis technology of deep learning to obtain a key parameter, integrating the key parameter and the quality evaluation result of the solid dye in the melting process, and generating a melting monitoring report.

The technical scheme of the application has the following beneficial effects:

The application generates an initial optical path through a multi-wavelength light source and an optical sensor, acquires an accurate time-light intensity curve by wavelet transformation denoising, accurately positions the melting starting moment by using a support vector machine model, calculates the melting rate by using a Kalman filter, predicts the stability by combining a thermodynamic analysis algorithm, and ensures the accurate monitoring of the melting process. When the melting rate reaches a preset threshold, automatically triggering a cooling program, analyzing the uniformity and purity of the coloring agent through a deep learning technology, and finally generating a detailed melting monitoring report. The method not only improves the monitoring precision, but also optimizes the production process, enhances the quality control capability of the product, and has obvious technical advantages and application value.

Further, the application provides a method for accurately identifying the moment when the solid dye starts to melt based on the change trend of the time-light intensity curve. Firstly, smoothing a time-light intensity curve by using a five-point three-time smoothing algorithm, calculating a slope value by combining a central difference method with a least square fitting technology, then, performing dimension reduction on the slope value by adopting principal component analysis, inputting the dimension reduced slope value into a support vector machine model trained in advance, classifying and identifying initial turning points, then, matching the curve in a specified time period by using a dynamic time warping algorithm based on the initial turning points, monitoring the model output by using an anomaly detection algorithm to predict and accurately position the actual turning points, and finally, optimizing the estimation of the melting start time by using a Kalman filter to generate a melting start signal.

By the method, various advanced algorithms including smoothing, slope calculation, dimension reduction, support vector machine classification, dynamic time warping, kalman filter optimization and the like are integrated, so that the high-precision identification of the melting starting point of the solid colorant is realized. Compared with the traditional method, the technology not only improves the accuracy of monitoring, but also enhances the adaptability to the complex nonlinear melting process, can effectively reduce misjudgment and noise interference, ensures accurate estimation of the melting starting moment, and further improves the product quality and the production efficiency. In addition, the method provides reliable data support for subsequent automatic control, and further optimizes the production process flow.

These and other aspects of the application will be more readily apparent from the following description of the embodiments.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 shows a flow chart of a melt monitoring method for solid colorants provided by the present application;

FIG. 2 shows a schematic diagram of a melt monitoring system for solid colorants according to the present application;

FIG. 3 illustrates a schematic diagram of a computing device provided by the present application.

Detailed Description

In order to enable those skilled in the art to better understand the present application, the following description will make clear and complete descriptions of the technical solutions according to the embodiments of the present application with reference to the accompanying drawings.

In some of the flows described in the specification and claims of the present application and in the foregoing figures, a plurality of operations occurring in a particular order are included, but it should be understood that the operations may be performed out of order or performed in parallel, with the order of operations such as 101, 102, etc., being merely used to distinguish between the various operations, the order of the operations themselves not representing any order of execution. In addition, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first" and "second" herein are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, and are not limited to the "first" and the "second" being different types.

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to fall within the scope of the application.

FIG. 1 is a flow chart of a method for monitoring the melting of solid colorants according to an embodiment of the present application, as shown in FIG. 1, the method comprising:

101. Placing a solid dye to be monitored in a transparent heating container, arranging at least one light source on one side of the transparent heating container, arranging at least one light sensor on the opposite side, selecting a wavelength combination according to the absorption spectrum of the solid dye, and utilizing a multi-wavelength light source to enable a light path to penetrate through the maximum section of the solid dye to generate an initial optical path;

Wherein the light sensor is a device capable of detecting and quantifying the intensity of light transmitted through the object, in the present system it is used to monitor the change in intensity of light transmitted through the solid dye, thereby providing data on the state of the dye;

Absorption spectrum, which is to represent the absorption degree of substances to light rays with different wavelengths, and the wavelength combination which is most suitable for monitoring specific substances (such as solid coloring agents) can be selected by analyzing the absorption spectrum;

The multi-wavelength light source is a light source capable of emitting light with various wavelengths, and the state change of the coloring agent can be reflected more accurately by selecting proper wavelength combinations, so that the monitoring precision is improved;

The initial optical path, the optical path formed by the multi-wavelength light source through the largest cross section of the solid dye, contains key information of the state of the dye and is the basis for subsequent analysis.

In actual operation, the solid dye to be monitored is placed in a transparent heating container, and a light source and a light sensor are respectively arranged at two sides of the solid dye. The appropriate combination of wavelengths is selected according to the absorbance spectrum of the stain, and the optical path is made to traverse the largest cross-section of the stain using a multi-wavelength light source, creating an initial optical path. This step provides the basis data for the subsequent time-light intensity curve.

For example, in an automated production line, the system automatically recognizes and loads a specified type of solid colorant into a transparent heated container. Then, the system configures a multi-wavelength light source according to a preset absorption spectrum, and starts the light source to irradiate the dye. At the same time, the light sensor starts to collect transmitted light intensity values in real time, and an initial optical path is initially constructed. The process ensures the accuracy of data acquisition and lays a foundation for the subsequent steps.

102. Recording the light intensity value received by the light sensor and transmitted through the solid dye according to a preset time interval to form an initial time-light intensity curve, and denoising the data of the initial optical path by a wavelet transformation algorithm to obtain a time-light intensity curve;

The time-light intensity curve shows the light intensity value changing along with time and reflects the dynamic change trend of the gradual melting of the solid coloring agent in the heating process;

The wavelet transformation algorithm is a signal processing technology, can effectively remove noise interference, simultaneously retains important characteristics of an original signal, and is suitable for processing complex time sequence data.

In actual operation, the light intensity values received by the light sensor are recorded according to preset time intervals, and an initial time-light intensity curve is formed. And then denoising the curve by using a wavelet transformation algorithm to obtain a smoother and more accurate time-light intensity curve so as to facilitate subsequent analysis.

Based on the above embodiment, the system records the light intensity values at a frequency of once per second, generating a time-light intensity curve. And then, denoising the curve by using a wavelet transformation algorithm, and eliminating the influence of random noise. The processed data is stored and ready for further analysis, ensuring the quality and reliability of the data.

103. Based on the change trend of the time-light intensity curve, observing by using a preset support vector machine model to obtain a turning point of which the slope of the light intensity curve changes, and indicating the moment when the solid coloring agent starts to melt based on the turning point to generate a melting start signal;

In the application scene, the support vector machine model is used for identifying turning points with the slope changing on a time-light intensity curve, namely the moment when the solid coloring agent starts to melt;

the melting start signal, which marks the precise moment when the solid colorant starts to melt, provides a key reference point for subsequent rate calculation and quality assessment.

In actual operation, based on the change trend of the time-light intensity curve, the slope change is observed by using a pre-trained support vector machine model, an initial turning point is determined, the moment when the solid dye starts to melt is indicated, and a melting start signal is generated.

Continuing with the above embodiment, the system uses a support vector machine model to analyze the denoised time-light intensity curve to identify turning points where the slope varies significantly. Once this turning point is determined, the system immediately generates a melt initiation signal and communicates it to the control system, ready for subsequent rate calculation and cooling procedures.

104. Calculating a melting rate by using a Kalman filter based on the data of a time-light intensity curve of the melting start signal in a fixed time, and predicting the stability of the solid coloring agent under different environmental conditions by combining a thermodynamic analysis algorithm to generate a quality evaluation report;

wherein the Kalman filter is a recursive filter for estimating a state variable of the system from a series of incomplete and noisy measurement data, which in this application is used to optimize the estimation of the melting rate;

The thermodynamic analysis algorithm predicts the stability of the solid coloring agent under different environmental conditions based on thermodynamic principles and helps to evaluate the product quality;

and a quality evaluation report, which is a report generated after comprehensively analyzing factors such as the melting rate, the environmental conditions and the like, and is used for evaluating the quality and the stability of the solid coloring agent.

In actual operation, the melting rate is calculated using a kalman filter on the data of the time-light intensity curve at a fixed timing based on the melting start signal. And predicting the stability of the solid coloring agent under different environmental conditions by combining a thermodynamic analysis algorithm, and generating a quality evaluation report.

According to the above embodiment, upon receiving the melting start signal, the system calculates the melting rate using a kalman filter and predicts the stability of the dye under different environmental conditions in combination with a thermodynamic analysis algorithm. The system automatically generates a detailed quality assessment report containing key parameters such as melting rate, temperature change and the like for reference by production management personnel to optimize the production process.

105. When the melting rate is detected to reach a preset threshold, a PID controller is utilized to automatically trigger a cooling program, the uniformity and purity of the solid coloring agent are predicted and analyzed by combining a deep learning image analysis technology, key parameters are obtained, key parameters and quality evaluation results of the solid coloring agent in the melting process are integrated, and a melting monitoring report is generated.

Wherein the PID controller is a proportional-integral-derivative controller for automatically adjusting a system parameter, such as temperature or speed, to achieve a set target value, in this application it is used to control the cooling program;

the deep learning image analysis technology is that a deep neural network is utilized to analyze the image, and the uniformity and purity of the coloring agent are identified and quantified;

and (3) integrating final reports of all key parameters and quality evaluation results in the melting process, and comprehensively knowing the whole melting process.

In actual operation, the PID controller is used to automatically trigger the cooling process when it is detected that the melt rate reaches a preset threshold. And predicting and analyzing the uniformity and purity of the solid coloring agent by combining the image analysis technology of deep learning to obtain key parameters. And integrating all key parameters and quality evaluation results of the coloring agent in the melting process to generate a melting monitoring report.

Continuing with the above embodiment, when the system detects that the melt rate reaches the preset threshold, the PID controller automatically initiates the cooling process to prevent product quality problems caused by overheating. Meanwhile, the system utilizes the deep learning technology to analyze the uniformity and purity of the coloring agent, and ensures the quality of the product. Finally, the system integrates all data to generate detailed melt monitoring reports for further analysis and improvement by production personnel.

Through the implementation of steps 101 to 105, the method realizes high-precision monitoring and intelligent control of the solid dye melting process by utilizing integrated light sensors, multi-wavelength light sources, a support vector machine model, a Kalman filter and a deep learning image analysis technology. The method not only improves the accuracy of monitoring, but also enhances the adaptability to the complex nonlinear melting process, reduces misjudgment and noise interference, and ensures accurate estimation of the melting starting time, rate and end point. In addition, the method provides reliable data support for subsequent automatic control, further optimizes the production process flow and improves the product consistency and the production efficiency.

In order to solve the problems of low accuracy and high misjudgment rate in the conventional method for identifying the melting start point of the solid colorant, further improve the accuracy and reliability of the melting monitoring, in some embodiments, in step 103, the step 103 of observing by using a preset support vector machine model based on the trend of the time-light intensity curve to obtain a turning point where the slope of the light intensity curve changes, and based on the turning point, indicating the time when the solid colorant starts to melt to generate a melting start signal, including:

The method comprises the steps of carrying out smoothing treatment on a time-light intensity curve by using a five-point three-time smoothing algorithm to obtain a smooth curve, calculating the slope of the smooth curve according to a central difference method and a least square fitting technology to obtain slope values corresponding to all time points, carrying out dimension reduction treatment on the slope values by using principal component analysis to obtain dimension reduction slope values, inputting the dimension reduction slope values into a support vector machine model trained in advance, classifying and identifying the dimension reduction slope values to obtain an initial turning point of change of the light intensity curve slope, matching the time-light intensity curve within a specified time period by using a dynamic time warping algorithm based on the initial turning point, monitoring model output by using an anomaly detection algorithm, predicting and positioning the turning point of change of the light intensity curve slope, and optimizing estimation of melting starting time by using a Kalman filter based on the turning point to generate a melting starting signal.

In the embodiment, a five-point three-time smoothing algorithm is a data smoothing technology, which reduces the influence of noise by using a weighted average value of five adjacent data points and is suitable for smoothing time series data;

The central difference method is a numerical differentiation method, is used for calculating the derivative (namely the slope) of discrete data points, and is used for estimating the slope of each point of a time-light intensity curve in the application scene so as to help identify key change points in the melting process;

A least squares fitting technique, a statistical method for determining a best fit straight line or curve by minimizing the sum of squares of errors between observed data and model predictions, is often used to extract useful feature information from noisy data;

A dynamic time warping algorithm (DTW), an algorithm for measuring the similarity between two time sequences, even though the sequences may have different speeds or lengths, in this scenario, the DTW is used to match the time-light intensity curve over a specified period of time, ensuring the accuracy of the turning point location;

The anomaly detection algorithm is a technology for identifying data points which do not accord with the expected mode in the data set, and is used for monitoring and supporting the quality of the output of the vector machine model in the application scene so as to improve the reliability of turning point positioning.

In the embodiment of the application, firstly, a five-point three-time smoothing algorithm is utilized to carry out smoothing treatment on a time-light intensity curve, noise interference is eliminated, and then a central difference method is applied to calculate the slope value corresponding to each time point by combining a least square fitting technology.

Then, the slope value is subjected to dimension reduction processing by adopting Principal Component Analysis (PCA) to obtain a dimension reduction slope value, and the dimension reduction slope value is input into a pre-trained support vector machine model for classification and identification so as to find a preliminary turning point. Based on the preliminary turning points, a dynamic time warping algorithm is used for matching curves in a specified time period, an anomaly detection algorithm is used for monitoring model output, and actual turning points are accurately predicted and positioned.

And finally, optimizing estimation of the melting start time by using a Kalman filter based on the positioned turning point to generate a melting start signal.

The following is a specific example:

For example, in an automated stain production line, the system first collects time-light intensity curve data for solid stain. In order to improve the data quality, the system applies a five-point three-time smoothing algorithm to carry out smoothing processing on the original curve, and the influence of random noise is obviously reduced. The system then uses a center difference method in combination with a least squares fitting technique to calculate the slope value for each point, identifying the key regions of slope variation.

Then, the system inputs the slope values into a pre-trained support vector machine model, and the model can more efficiently identify the initial turning point of the slope change after the principal component analysis and dimension reduction processing. Based on the preliminary turning point, the system further applies a dynamic time warping algorithm to match the curve within a specified time period, so as to ensure the accuracy of turning point positioning. Meanwhile, an anomaly detection algorithm is used for monitoring the model output, so that the reliability of the identification result is ensured.

Once the system determines the final turning point, i.e. the moment at which the solid colorant starts to melt, a kalman filter is used to optimize the estimate of this moment, generating an accurate melt initiation signal. This signal not only provides the basis for subsequent rate calculations, but also triggers a series of automated control procedures, such as automatically adjusting the heating power or starting the cooling procedure, thus achieving a highly automated and intelligent management of the whole production process.

By the method for comprehensively applying various advanced algorithms and technologies, the system not only improves the accuracy and reliability of melting monitoring, but also enhances the adaptability to complex nonlinear melting processes, and remarkably improves the product quality and production efficiency.

In order to solve the complexity problem of the high-dimensional slope value matrix in processing and analysis, further improve accuracy and efficiency of classification of the support vector machine model, in some embodiments, the method for performing the dimension reduction processing on the slope value by using principal component analysis to obtain a dimension reduction slope value includes:

optimizing the slope values by using a robust regression algorithm, organizing to form a high-dimensional slope value matrix, standardizing the characteristics of the high-dimensional slope value matrix according to a Z-Score standardization method to generate a standardized slope value matrix, wherein the characteristics comprise zero mean and unit variance, calculating a predefined covariance matrix based on a Mah distance according to the standardized slope value matrix, accelerating the decomposition process of the covariance matrix by using a random singular value decomposition algorithm to obtain characteristic values and characteristic vectors, evaluating the performances of the support vector machine model under different principal component numbers by using a cross verification technology based on the characteristic values and the characteristic vectors, selecting N principal components which can retain maximized information and verify the performance of the support vector machine model, adopting minimum absolute shrinkage and a selection operator to carry out regression screening on principal components which are helpful to classification tasks when determining the N values, generating a principal component set, carrying out preliminary dimension reduction on the standardized slope value matrix by using a t-distribution random neighborhood algorithm to obtain a preliminary dimension reduction slope value, and projecting the dimension reduction slope value onto the principal component defined by the principal component to generate a new dimension value set.

In this embodiment, a robust regression algorithm is a regression analysis method for processing the presence of outliers in a dataset; in the application scene, a robust regression algorithm is used for optimizing the slope value so as to improve the reliability of subsequent analysis;

The high-dimensional slope value matrix is a data matrix formed by slope values corresponding to a plurality of time points and generally has higher dimension, and comprises key information of different time points in the solid coloring agent melting process;

The Z-Score standardization method is a data standardization technology, which converts data into the form of zero mean and unit variance, is helpful to eliminate the dimension difference between different features, and is convenient for subsequent statistical analysis and machine learning model training;

in the application scene, the covariance matrix is calculated based on the Markov distance, so that the correlation between the data points can be captured more accurately;

the Random Singular Value Decomposition (RSVD) is an algorithm for accelerating large matrix decomposition, and by the random sampling technology, the RSVD can greatly reduce the computational complexity while keeping higher precision, and is suitable for large-scale data processing;

Cross-validation techniques, a method of evaluating the performance of a machine learning model by dividing a dataset into subsets for multiple training and testing to select optimal model parameters or structures, which in this application are used to determine optimal values for the number of principal components;

a regularized regression method, which compresses unimportant regression coefficients to zero by introducing penalty terms so as to realize feature selection; in this scenario, LASSO is used to screen out principal components that are helpful to classification tasks;

a t distribution random neighborhood embedding (t-SNE) is a dimension reduction technology which is particularly suitable for visualization and cluster analysis of high-dimension data, and is mapped into a low-dimension space by simulating probability distribution in the high-dimension space, so that a local structure is reserved.

In the embodiment of the application, firstly, a robust regression algorithm is utilized to optimize the slope value, and a high-dimensional slope value matrix is formed by organization. And then, performing standardization processing on the high-dimensional slope value matrix by adopting a Z-Score standardization method to generate a standardized slope value matrix. And then, calculating a covariance matrix based on the Markov distance according to the standardized slope value matrix, and accelerating the decomposition process by using a random singular value decomposition algorithm to obtain a characteristic value and a characteristic vector.

Next, the performance of the support vector machine model is evaluated based on cross-validation techniques for different numbers of principal components, N principal components are selected that maximize retention information and validate model performance, while principal components that contribute to classification tasks are screened out using LASSO, generating a set of principal components.

Finally, the normalized slope value matrix is subjected to preliminary dimension reduction by using a t-SNE algorithm, and the preliminary dimension reduction slope value is projected onto a new coordinate system defined by a main component set to generate a final dimension reduction slope value.

The following is a specific example:

According to the above embodiment, the system first collects time-light intensity curve data during the melting process, and calculates the slope value at each time point to form a high-dimensional slope value matrix. In order to improve the quality and stability of data, the system adopts a robust regression algorithm to optimize the slope value, so that the influence of abnormal values is reduced.

The system then normalizes the optimized slope values to zero mean and unit variance form, generating a normalized slope value matrix. Based on the standardized matrix, the system calculates a covariance matrix based on the mahalanobis distance, and rapidly decomposes the matrix by using a random singular value decomposition algorithm to extract the eigenvalues and eigenvectors.

In order to determine the optimal principal component quantity, the system adopts a cross-validation technology to evaluate the performance of the support vector machine model under different principal component quantities, and simultaneously utilizes LASSO regression to screen out principal components which are most helpful to classification tasks, so as to generate a principal component set. The process not only improves the accuracy of the model, but also ensures the generalization capability of the model.

Next, the system performs preliminary dimension reduction on the standardized slope value matrix by using a t-SNE algorithm, and projects the slope value after preliminary dimension reduction onto a new coordinate system defined by the main component set to generate a dimension reduction slope value. The dimension-reduced data not only reduces the computational complexity, but also reserves the key characteristics of the original data, and provides high-quality input data for the subsequent support vector machine model classification.

By the method for comprehensively applying a plurality of advanced algorithms and technologies, the system not only improves the precision and the efficiency of data processing, but also enhances the adaptability to complex nonlinear melting processes, and remarkably improves the product quality and the production efficiency. The method provides powerful support for intelligent management of an automatic control system.

In order to solve the problem that the uniformity and purity of the solid colorant are difficult to control accurately in the melting process, the product quality and the production efficiency are further improved, in some embodiments, when the melting rate is detected to reach the preset threshold value in step 105, a PID controller is utilized to automatically trigger a cooling program, the uniformity and purity of the solid colorant are predicted and analyzed by combining the deep learning image analysis technology, key parameters are obtained, the key parameters and the quality evaluation result of the solid colorant in the melting process are integrated, and a melting monitoring report is generated, which comprises:

When the melting rate reaches a preset threshold value, a parameter setting of a PID controller is optimized by utilizing fuzzy logic, a cooling program is automatically triggered according to a preset temperature control strategy, in the cooling process, an initial image of the solid coloring agent is obtained by using a high-resolution camera, the initial image is processed based on a multi-scale image fusion technology to obtain an optimized image, the optimized image is preprocessed according to a super-pixel segmentation algorithm to obtain a processed image, characteristics such as color distribution, particle size and distribution extracted from the processed image are analyzed by utilizing an integrated learning method to obtain key parameters including uniformity index, purity grade and impurity content of the solid coloring agent, external data in the melting process are collected according to the key parameters, melting rate and environmental condition change in future fixed time are predicted based on time sequence analysis to obtain an analysis result, the external data comprises the melting rate, the environmental temperature and the humidity, and a natural language processing technology is applied to automatically generate a melting monitoring report based on the analysis result.

In this embodiment, the fuzzy logic optimizes the PID controller, a fuzzy logic-based control strategy for optimizing the parameter settings of the PID (proportional-integral-derivative) controller, which processes uncertainty and inaccurate information by mimicking the human decision process, thereby achieving more flexible and robust control;

The method can improve the detail definition and the whole quality of the image, and is particularly suitable for complex scene images acquired by a high-resolution camera;

The super-pixel segmentation algorithm is an image segmentation technology for simplifying image representation by grouping pixels into super-pixels, wherein each super-pixel consists of a group of adjacent pixels with similar color or texture characteristics, so that the subsequent feature extraction and analysis are facilitated;

The common integrated learning method comprises random forests, gradient lifting trees and the like, and is used for extracting and analyzing key features from processed images in the application;

Time series analysis, a statistical method for analyzing time-varying data sequences to identify trends, periodicity, and other patterns, in this application scenario, it is used to predict future melt rate and environmental condition changes;

natural Language Processing (NLP), a field of computer science, relates to techniques that enable computers to understand, interpret and generate human language, and in this application scenario, NLP is used to automatically generate detailed melt monitoring reports.

In the embodiment of the application, when the melting rate of the solid dye is detected to reach the preset threshold, the system optimizes the parameter setting of the PID controller by using the fuzzy logic, and automatically triggers a cooling program according to a preset temperature control strategy. During the cooling process, an initial image of the solid colorant is captured using a high resolution camera and processed using a multi-scale image fusion technique to obtain an optimized image. Then, the optimized image is preprocessed by applying a super-pixel segmentation algorithm to generate a processed image. And then, extracting the characteristics of color distribution, particle size, distribution and the like from the processed image by using an ensemble learning method, and calculating key parameters such as uniformity index, purity grade, impurity content and the like of the solid coloring agent. Based on these key parameters, the system collects external data during the melting process (such as melt rate, ambient temperature and humidity) and predicts melt rate and ambient condition changes over a fixed time in the future by time series analysis. Finally, natural language processing technology is applied to automatically generate a melting monitoring report containing all key parameters and quality evaluation results.

The following is a specific example:

Continuing with the above embodiment, when the system detects that the melt rate reaches a preset threshold, the fuzzy logic optimized PID controller automatically adjusts its parameter settings and initiates a cooling procedure. To ensure cooling effect and product quality, the system captures a status image of the solid colorant in real time using a high resolution camera.

These initial images are first processed through a multi-scale image fusion technique to enhance the details and sharpness of the image, generating an optimized image. Next, the system applies a super-pixel segmentation algorithm to preprocess the optimized image, decomposing the image into a plurality of super-pixel regions, facilitating subsequent feature extraction.

On the basis of processing the image, the system analyzes the characteristics of color distribution, particle size, distribution and the like extracted from the image by using an ensemble learning method, and calculates key parameters such as uniformity index, purity grade, impurity content and the like of the solid coloring agent. For example, the uniformity of the stain can be quantified by analyzing the color differences in the different regions, and the purity and impurity content can be assessed by measuring the particle size and distribution.

At the same time, the system collects external data during the melting process (such as melt rate, ambient temperature and humidity) and uses time series analysis methods to predict changes in melt rate and ambient conditions over a period of time in the future. This step not only helps the producer prepare in advance, but also provides important feedback information for optimizing the production flow.

Finally, the system automatically generates detailed melt monitoring reports using natural language processing techniques. The report contains all key parameters, quality assessment results and predictions of future trends, and provides comprehensive support for production management and quality control. The scheme comprehensively utilizes a plurality of advanced technologies and methods, and remarkably improves the automation level and the product quality of the solid coloring agent production.

In order to solve the problems of low precision and edge information loss in the processing of solid dye images in the conventional image segmentation method, the accuracy and detail retention of image segmentation are further improved, in some embodiments, the optimized image is preprocessed according to a super-pixel segmentation algorithm, so as to obtain a processed image, which includes:

The method comprises the steps of obtaining a solid dye, carrying out weighted average fusion processing on an initial image of the solid dye obtained by a high-resolution camera by utilizing a multi-view data set and a structural similarity index to generate an optimized image, initializing a super-pixel grid according to the optimized image, carrying out iterative optimization processing on the boundary of the super-pixel grid based on an improved simple linear iterative clustering algorithm and a spectral clustering technology to obtain a preliminary segmentation result, carrying out refinement adjustment and smoothing processing on the boundary of the super-pixel grid based on the preliminary segmentation result by applying an active contour model and a graph cutting algorithm, simultaneously introducing shape regularity constraint such as ellipse fitting to generate an intermediate segmentation image, removing isolated areas according to the intermediate segmentation image by adopting a connected component marking algorithm, and carrying out morphological operation by applying bilateral filtering and smoothing operation to generate a processed image, wherein the morphological operation comprises open operation and closed operation.

In this embodiment, the multi-view dataset is a dataset made up of images taken from a plurality of different views, capable of providing more comprehensive target information;

A Structural Similarity Index (SSIM), which is a measure of similarity between two images, taking into account the differences in brightness, contrast and structure, in this scenario, SSIM is used to guide the weighted average fusion process to generate a high quality optimized image;

a super-pixel grid for dividing the image into a grid structure composed of a plurality of super-pixels, each super-pixel being composed of a group of adjacent pixels having similar color or texture characteristics;

the simple linear iterative clustering algorithm (SLIC) is a common super-pixel segmentation algorithm, and simplifies image representation by dividing an image into a fixed number of super-pixels;

The spectral clustering technology is a clustering method based on graph theory, and performs clustering analysis by constructing a similarity matrix among image pixels, and can effectively capture complex structural features in images;

an active contour Model (Snake Model) is an image segmentation method that gradually approximates the curve to the target boundary by minimizing the energy function, which in this application is used to refine the boundary of the super-pixel grid;

The Graph Cut algorithm (Graph Cut) is an image segmentation method based on Graph theory, which is used for segmenting an image by searching a minimum Cut, and is commonly used for refining and optimizing a segmentation result;

Shape regularity constraints (e.g., ellipse fitting), a method for constraining the shape of the segmentation result, ensuring that the segmentation region meets specific geometric requirements;

the communication component marking algorithm is an algorithm for identifying and marking the communication region in the image, and can remove isolated small regions and improve the accuracy of a segmentation result;

Bilateral filtering, namely a nonlinear filtering method which can smooth image noise while maintaining edges, wherein the nonlinear filtering method is used for preprocessing before morphological operation in the application scene;

Morphological operations (open and closed operations) a series of image processing operations based on set theory for removing small objects or filling holes, open operations are typically used to remove fine noise points, and closed operations are used to fill holes.

In the embodiment of the application, firstly, the initial image acquired by the high-resolution camera is subjected to weighted average fusion processing by utilizing the multi-view data set and the structural similarity index to generate an optimized image. And initializing the super-pixel grid according to the optimized image, and carrying out iterative optimization processing on the boundary of the super-pixel grid by adopting a modified version of simple linear iterative clustering algorithm and a spectral clustering technology to obtain a preliminary segmentation result.

And then, based on the preliminary segmentation result, applying an active contour model and a graph cutting algorithm to carry out refinement adjustment and smoothing treatment on the boundary of the super-pixel grid, and simultaneously introducing shape regularity constraint (such as ellipse fitting) to generate an intermediate segmentation image. Finally, the isolated area is removed by using a connected component marking algorithm according to the intermediate segmentation image, and morphological operation (comprising open operation and closed operation) is carried out by applying bilateral filtering and smoothing operation, so as to generate a final processing image. The following is a specific example:

Continuing with the above embodiment, the system captures images of the solid colorant from multiple angles using a high resolution camera, forming a multi-view dataset. To generate high quality optimized images, the system performs a weighted average fusion process on the images using the structural similarity index. This step not only improves the sharpness of the image, but also enhances the detailed information in the image.

Then, the system initializes the super-pixel grid based on the optimized image, and adopts a modified version of simple linear iterative clustering algorithm to carry out iterative optimization processing on the boundary of the super-pixel grid in combination with a spectral clustering technology. The method can accurately capture the change of the microstructure in the dye and generate a preliminary segmentation result.

To further refine and optimize the segmentation results, the system applies an active contour model and a graph cut algorithm to adjust and smooth the boundaries of the superpixel mesh and introduce shape regularity constraints (e.g., ellipse fitting). These steps ensure that the boundaries of the segmented regions are smoother and more regular, resulting in an intermediate segmented image.

The system then uses a connected component labeling algorithm to remove isolated regions in the intermediate segmented image, avoiding these small regions from interfering with subsequent analysis. To further improve image quality, the system performs morphological operations (including open and closed operations) using bilateral filtering and smoothing operations to generate a final processed image.

By comprehensively applying the scheme of a plurality of advanced image processing technologies and methods, the system not only improves the precision and reliability of image segmentation, but also enhances the adaptability to complex nonlinear melting process, and remarkably improves the product quality and production efficiency. The method provides powerful support for intelligent management and optimization of an automatic control system.

To further improve the accuracy and reliability of the melt rate calculation in order to solve the problems of noise interference and outlier effects during the melt rate calculation, in some embodiments, the calculating the melt rate using a kalman filter based on the data of the time-light intensity curve of the melt initiation signal at a fixed time in step 104 includes:

The method comprises the steps of utilizing a melting start signal, combining an adaptive threshold algorithm, calculating a fixed moment when a solid coloring agent starts to melt, adopting a dynamic window selection method based on the fixed moment, automatically adjusting the time period length according to the change trend of melting the melting rate, extracting initial time-light intensity curve data starting from the fixed moment, utilizing a median filtering method to remove noise and abnormal values of the initial time-light intensity curve data, simultaneously introducing a local weighted regression scattered point smoothing technology to retain characteristic information in the initial time-light intensity curve data, further utilizing a robust statistical method to automatically detect and remove abnormal values to obtain smoothed time-light intensity curve data, utilizing a five-point cubic smoothing algorithm to reduce noise interference based on the smoothed time-light intensity curve data, utilizing a least square fitting technology to combine an adaptive weight adjustment strategy to calculate a slope value corresponding to each time point, representing the initial melting rate, and utilizing a unscented Kalman filter to optimize the initial melting rate, and obtaining the initial melting rate.

In the embodiment, the self-adaptive threshold algorithm is a method for dynamically adjusting the threshold value, and automatically adjusts the threshold level according to the real-time change of the data, wherein the self-adaptive threshold algorithm is used for determining the accurate moment when the solid coloring agent starts to melt in the application scene;

The dynamic window selection method is a flexible data processing technology, and is beneficial to more accurately extracting and analyzing time-light intensity curve data by automatically adjusting the analyzed time period length according to the change trend of the data so as to better capture the data characteristics;

A data smoothing technique based on weighted average, which reduces noise interference by performing a polynomial fitting for three times on each data point and four adjacent points; the method can effectively remove high-frequency noise and retain the main trend of data;

The least square fitting technology is combined with an adaptive weight adjustment strategy, namely a statistical method, a best fit straight line or curve is determined by minimizing the sum of squares of errors between observed data and model predictions, and the fitting effect is optimized by combining with the adaptive weight adjustment strategy;

Unscented Kalman Filter (UKF), an improved version of the extended Kalman filter, is suitable for use in nonlinear systems, and the UKF approximates the probability distribution of the system state by using a carefully selected set of sample points (called Sigma points), thereby improving the accuracy of the state estimation.

In the embodiment of the application, firstly, a melting start signal is utilized and an adaptive threshold algorithm is combined to calculate the fixed moment when the solid coloring agent starts to melt. Based on this fixed time, a dynamic window selection method is adopted to automatically adjust the time period length according to the change trend of the melting rate, and the initial time-light intensity curve data from the time is extracted.

And then removing noise and abnormal values in the initial data by using a median filtering method, simultaneously introducing a local weighted regression scattered point smoothing technology to retain characteristic information, and further automatically detecting and removing abnormal values by using a robust statistical method to obtain smoothed time-light intensity curve data.

And then, based on the smoothed data, further reducing noise interference by using a five-point cubic smoothing algorithm, and calculating a slope value corresponding to each time point by using a least square fitting technology and combining an adaptive weight adjustment strategy, wherein the slope value is expressed as a primary melting rate.

Finally, the initial melting rate is optimized by using an unscented Kalman filter, and a more accurate melting rate is obtained.

The following is a specific example:

Continuing with the above embodiment, when the system detects a signal that solid colorant begins to melt, an adaptive threshold algorithm is first used to determine the exact start of melting. Based on this time, the system automatically adjusts the analysis period according to the trend of the melting rate by using a dynamic window selection method, and extracts the initial time-light intensity curve data from the melting start time.

To ensure the quality of the data, the system first removes noise and outliers in the initial time-light intensity curve data using a median filtering method. Next, a locally weighted regression scatter smoothing technique is applied to preserve key feature information in the data and further a robust statistical method is used to automatically detect and remove any remaining outliers, generating smoothed time-light intensity curve data.

After obtaining high-quality smooth data, the system further reduces noise interference by applying a five-point cubic smoothing algorithm, and calculates the slope value of each time point by combining a least square fitting technology with an adaptive weight adjustment strategy to serve as a primary melting rate. This process not only improves the accuracy of the data, but also enhances the adaptability to complex nonlinear melting processes.

Finally, the system optimizes the preliminary melting rate by using an unscented Kalman filter to obtain a more accurate melting rate. The optimized melting rate not only provides more reliable monitoring results, but also provides important basis for subsequent cooling procedures and other automatic control operations. By the scheme of comprehensively applying a plurality of advanced technologies and methods, the system remarkably improves the accuracy and reliability of melting monitoring, further optimizes the production process flow and improves the product quality and production efficiency.

In order to solve the problem that noise and abnormal values in the initial time-light intensity curve data affect analysis precision, further improve the effect of data smoothing and retention of characteristic information, in some embodiments, a median filtering method is used to remove noise and abnormal values in the initial time-light intensity curve data, and meanwhile a local weighted regression scattered point smoothing technology is introduced to retain characteristic information in the initial time-light intensity curve data, and a robust statistical method is further applied to automatically detect and remove abnormal values to obtain smoothed time-light intensity curve data, including:

The method comprises the steps of carrying out preliminary denoising on initial time-light intensity curve data by adopting a median filtering method, reducing noise influence by replacing each data point with a median in an adjacent area to obtain preliminary denoised data, smoothing a curve by carrying out weighted least square fitting on local data points based on the preliminary denoised data and introducing a local weighted regression scattered point smoothing technology, retaining characteristic information and local variation trend to generate smoothed curve data, automatically detecting and removing abnormal values according to the smoothed curve data by adopting a robust statistical method to obtain a detection result, and integrating the smoothed time-light intensity curve data based on the detection result, the preliminary denoised data and the smoothed curve data.

In this embodiment, the median filtering method, a nonlinear digital filtering technique, reduces noise by replacing each data point with the median of the data points in the neighborhood, is particularly useful for removing impulse noise (also known as salt and pepper noise) while preserving as much as possible the main characteristics of the original signal;

A local weighted regression scattered point smoothing technology (LOESS) is a non-parametric regression method for smoothing a curve by carrying out weighted least square fitting on local data points, and can effectively capture local variation trend in data and retain important characteristic information at the same time;

Robust statistical methods, which are statistical techniques that can still provide reliable results in the presence of outliers or noise, include RANSAC (random sample consensus algorithm), M estimation, etc., which can automatically detect and remove outliers in data, thereby improving the accuracy of analysis results.

In the embodiment of the application, firstly, a median filtering method is adopted to perform preliminary denoising treatment on initial time-light intensity curve data, and each data point is replaced by a median in the neighborhood of the data point to reduce noise influence, so that preliminary denoised data is generated.

And then, based on the data after preliminary denoising, introducing a local weighted regression scattered point smoothing technology (LOESS), smoothing a curve by carrying out weighted least square fitting on local data points, retaining characteristic information and local variation trend, and generating smoothed curve data.

And then, automatically detecting and removing abnormal values in the smoothed curve data by using a robust statistical method to obtain a detection result. And finally, integrating and generating final smooth time-light intensity curve data based on the detection result, the data after preliminary denoising and the smoothed curve data.

The following is a specific example:

Continuing with the above example, the system collects time-light intensity profile data for solid colorants during the melting process. To ensure the quality of these data, the initial data is first subjected to a preliminary denoising process using a median filtering method. Specifically, for each data point, the system will calculate the median value in its neighborhood and replace the original data point with this median value. This effectively removes random noise, especially impulse noise, while retaining the main trend information.

Next, the system introduces a locally weighted regression-scattered-point smoothing technique (LOESS) based on the preliminary denoised data. By assigning weights to the local regions near each data point and performing a weighted least squares fit, the system is able to smooth the curve while preserving significant local variation trends and characteristic information. For example, within certain critical time periods, the light intensity of the stain may vary significantly, and these trends need to be accurately captured and retained.

The system then automatically detects and removes outliers in the smoothed curve data using a robust statistical method (e.g., RANSAC). These outliers may be due to sensor failures or other external disturbances, whose presence can affect the accuracy of subsequent analysis. By means of a robust statistical method, the system can identify and reject these outliers, ensuring the authenticity and reliability of the remaining data.

And finally, integrating the data after preliminary denoising, the smoothed curve data and the detection result of the robust statistical method by the system to generate final smoothed time-light intensity curve data. These high quality data not only provide a solid basis for subsequent melt rate calculations, but also enhance the automation control and intelligent management level of the overall production process. By the scheme of comprehensively applying a plurality of advanced technologies and methods, the system remarkably improves the accuracy and reliability of data processing, further optimizes the production process flow and improves the product quality and production efficiency.

FIG. 2 is a schematic diagram of a device (or system) for monitoring the melting of solid colorants according to an embodiment of the present application, as shown in FIG. 2, the device comprising:

A generating module 21, configured to place a solid dye to be monitored in a transparent heating container, set at least one light source on one side of the transparent heating container, set at least one light sensor on the opposite side, and select a wavelength combination according to an absorption spectrum of the solid dye, and make a light path pass through a maximum section of the solid dye by using a multi-wavelength light source to generate an initial optical path;

the denoising module 22 is configured to record the light intensity value received by the light sensor and transmitted through the solid dye according to a preset time interval, form an initial time-light intensity curve, and perform denoising processing on the data of the initial optical path by using a wavelet transformation algorithm to obtain a time-light intensity curve;

The indicating module 23 is configured to observe, based on the trend of the time-light intensity curve, with a preset support vector machine model, to obtain a turning point at which the slope of the light intensity curve changes, and instruct, based on the turning point, a time at which the solid colorant begins to melt, so as to generate a melting start signal;

an evaluation module 24, configured to calculate a melting rate using a kalman filter based on data of a time-light intensity curve of the melting start signal within a fixed time, predict stability of the solid colorant under different environmental conditions in combination with a thermodynamic analysis algorithm, and generate a quality evaluation report;

And the monitoring module 25 is used for automatically triggering a cooling program by utilizing a PID controller when the melting rate reaches a preset threshold value, predicting and analyzing the uniformity and purity of the solid coloring agent by combining a deep learning image analysis technology to obtain key parameters, and integrating the key parameters and quality evaluation results of the solid coloring agent in the melting process to generate a melting monitoring report.

A melting monitoring apparatus for solid colorants as shown in fig. 2 may perform a melting monitoring method for solid colorants as shown in the embodiment of fig. 1, and the principle and technical effects thereof will not be described again. The specific manner in which the individual modules, units, of the melt monitoring device for solid colorants in one of the above embodiments perform the operation has been described in detail in connection with the embodiments of the method and will not be described in detail herein.

In one possible design, a melt monitoring apparatus for solid colorants of the embodiment shown in FIG. 2 may be implemented as a computing device, as shown in FIG. 3, which may include a storage component 31 and a processing component 32;

The storage component 31 stores one or more computer instructions for execution by the processing component 32.

The processing assembly 32 is used in one of the melt monitoring methods for solid colorants described above with respect to the embodiment of FIG. 1.

Wherein the processing component 32 may include one or more processors to execute computer instructions to perform all or part of the steps of the methods described above. Of course, the processing component may also be implemented as one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic elements for executing the methods described above.

The storage component 31 is configured to store various types of data to support operations at the terminal. The memory component may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.

Of course, the computing device may necessarily include other components as well, such as input/output interfaces, display components, communication components, and the like.

The input/output interface provides an interface between the processing component and a peripheral interface module, which may be an output device, an input device, etc.

The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, and the like.

The computing device may be a physical device or an elastic computing host provided by the cloud computing platform, and at this time, the computing device may be a cloud server, and the processing component, the storage component, and the like may be a base server resource rented or purchased from the cloud computing platform.

The embodiment of the application also provides a computer storage medium storing a computer program which, when executed by a computer, can implement a XX method of the embodiment shown in fig. 1.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.

It should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same, and although the present application has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that the technical solution described in the above-mentioned embodiments may be modified or some technical features may be equivalently replaced, and these modifications or substitutions do not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solution of the embodiments of the present application.

Claims

1. A melt monitoring method for a solid colorant, comprising:

when the melting rate is detected to reach a preset threshold, automatically triggering a cooling program by using a PID controller, and predicting and analyzing the uniformity and purity of the solid coloring agent by combining a deep learning image analysis technology to obtain key parameters, integrating the key parameters and quality evaluation results of the solid coloring agent in the melting process, and generating a melting monitoring report;

When the melting rate is detected to reach a preset threshold, a PID controller is utilized to automatically trigger a cooling program, the uniformity and purity of the solid coloring agent are predicted and analyzed by combining a deep learning image analysis technology, key parameters are obtained, key parameters and a quality evaluation result of the solid coloring agent in the melting process are integrated, and a melting monitoring report is generated, and the method comprises the following steps:

based on the analysis result, automatically generating a melting monitoring report by applying a natural language processing technology;

Preprocessing the optimized image according to a super-pixel segmentation algorithm to obtain a processed image, including:

2. The method according to claim 1, wherein observing with a preset support vector machine model based on the trend of the time-light intensity curve to obtain a turning point at which the slope of the light intensity curve changes, and indicating the time at which the solid colorant starts to melt based on the turning point, to generate a melting start signal, comprises:

3. The method of claim 2, wherein the dimensionality reduction of the slope value using principal component analysis results in a dimensionality reduction slope value, comprising:

4. The method as set forth in claim 1, wherein calculating the melting rate using a kalman filter based on data of a time-light intensity curve of the melting start signal at a fixed time, includes:

5. The method of claim 4, wherein removing noise and outliers from the initial time-light intensity curve data using a median filtering method while preserving characteristic information in the initial time-light intensity curve data by introducing a locally weighted regression-scatter smoothing technique, and further automatically detecting and removing outliers using a robust statistical method, and obtaining smoothed time-light intensity curve data, comprises:

6. A melt monitoring system for a solid colorant, comprising:

The monitoring module is used for automatically triggering a cooling program by utilizing a PID controller when the melting rate reaches a preset threshold value, predicting and analyzing the uniformity and purity of the solid coloring agent by combining a deep learning image analysis technology to obtain key parameters, integrating the key parameters and quality evaluation results of the solid coloring agent in the melting process, and generating a melting monitoring report;

7. A computing device, comprising a processing component and a storage component, the storage component storing one or more computer instructions for execution by the processing component, the one or more computer instructions to implement a melt monitoring method for a solid colorant according to any one of claims 1-5.

8. A computer storage medium storing a computer program which, when executed by a computer, implements a melt monitoring method for a solid colorant according to any one of claims 1 to 5.