CN117612195A - Graph model generation method and device based on main wiring graph recognition technology - Google Patents
Graph model generation method and device based on main wiring graph recognition technology Download PDFInfo
- Publication number
- CN117612195A CN117612195A CN202311445433.3A CN202311445433A CN117612195A CN 117612195 A CN117612195 A CN 117612195A CN 202311445433 A CN202311445433 A CN 202311445433A CN 117612195 A CN117612195 A CN 117612195A
- Authority
- CN
- China
- Prior art keywords
- wiring diagram
- file
- identification
- model
- recognition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/42—Document-oriented image-based pattern recognition based on the type of document
- G06V30/422—Technical drawings; Geographical maps
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/148—Segmentation of character regions
- G06V30/15—Cutting or merging image elements, e.g. region growing, watershed or clustering-based techniques
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/19007—Matching; Proximity measures
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1916—Validation; Performance evaluation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/19167—Active pattern learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1918—Fusion techniques, i.e. combining data from various sources, e.g. sensor fusion
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A pattern generation method and device based on a main wiring diagram recognition technology. The method comprises the steps of transmitting CAD graph files in a substation monitoring system to an algorithm service rear end through a front-end tool, and identifying main wiring graph elements of the CAD graph files; comparing and analyzing the identification result with the CAD graph file, updating the identification result according to revision information input by a user, converting the result into a power graph file and a power model file, and returning the power graph file and the power model file to the substation monitoring system; the revision information is fed back to the algorithm service back end to update and train the identification model; and regenerating power model data according to the power pattern file stored by the transformer substation monitoring system, and transmitting the power model data and the power model file to the algorithm service back end for comparison and verification, and processing model differences. According to the scheme, the power pattern data and the power pattern data are automatically generated, so that the manual pressure is reduced, and the recognition accuracy is improved.
Description
Technical Field
The invention belongs to the field of power automation, and particularly relates to a pattern generation method and device based on a main wiring pattern recognition technology.
Background
The power dispatching control system adopts a power graph format file based on a general information model to describe a plant station picture. The electric power model format file describes an electric equipment model, a dispatching operation and maintenance personnel needs to draw an electric power graph picture and model the electric equipment by adopting a manual drawing and recording mode and referring to a plant station wiring diagram design original drawing. Because of the complex pattern and numerous equipment types, the maintenance work is complicated, and meanwhile, the problems of attribute deletion, association errors, connection line virtual connection and the like are extremely easy to occur. The identification method of the main wiring image in the prior art mainly comprises a corner detection algorithm, a mathematical morphology or a template matching method. The image is usually subjected to morphological processing to obtain an image matrix of various devices, so as to obtain feature vectors to realize identification or matching of different devices. According to the method, global searching and matching are needed on the image, but the calculation amount is large due to the fact that the image of the electric main wiring diagram of the transformer substation monitoring system is large, meanwhile, due to the fact that the drawing standard of the CAD drawing is inconsistent, the deformation of the graphic elements is large, the recognition inaccuracy is caused by the fact that characters and connecting lines in the image interfere with the recognition of the graphic elements, the algorithm can be changed and upgraded when the graphic element types are newly added, the problems of noise interference and the like exist in the scanned electronic image, the recognition accuracy rate of the main wiring diagram is low, the speed is low, and practical application is difficult.
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides a pattern generation method and device based on a main wiring diagram recognition technology, so as to solve the technical problem of automatic generation of power pattern data and power model data.
In order to solve the technical problems, the invention adopts the following technical scheme.
The invention firstly discloses a pattern generation method based on a main wiring diagram recognition technology, which comprises the following steps:
step 1, a CAD graph file to be analyzed in a transformer substation monitoring system is transmitted to an algorithm service rear end through a front end tool, and main wiring graph elements of the CAD graph file are identified based on an identification model of the algorithm service rear end, so that graphic element information, text information and topology information are obtained;
step 2, comparing and analyzing the main wiring diagram recognition result with the CAD graph file, visualizing the difference elements, updating the main wiring diagram recognition result according to revision information input by a user, converting the main wiring diagram recognition result into a power graph file and a power model file, and returning to the substation monitoring system;
step 3, feeding back the revised information to the algorithm service back end, and updating and training the identification model;
and 4, regenerating power model data according to the power pattern file stored by the transformer substation monitoring system, and comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end to process model differences.
The invention further comprises the following preferable schemes:
in the step 1, a CAD graphics file to be analyzed in the substation monitoring system is transmitted to the algorithm service back end through a front end tool, and the method further includes:
the front-end tool is deployed in the transformer substation monitoring system, the transformer substation monitoring system and the front-end tool communicate through process calling, files are transmitted to the front-end tool through calling parameters, the front-end tool acquires CAD graphic files, analysis tasks and analysis data are transmitted to the algorithm service rear end through a message queue mode to carry out asynchronous analysis calling, and the algorithm service rear end operates on an independent algorithm server.
Identifying the main wiring diagram element of the CAD graph file to obtain primitive information, text information and topology information, and further comprising:
step 1.1: carrying out deformation and enhancement treatment on the training sample, adopting an identification algorithm based on an image segmentation pyramid to realize the identification of the ultra-high resolution main wiring diagram, fusing sub-graph identification results under each scale, and further screening and de-duplicating the multi-scale identification results to obtain a final identification result;
step 1.2: performing transfer learning and customized training on the OCR recognition model, performing mutual verification and association by combining a topological relation and primitive elements, deleting pictures of equipment primitives, reducing the number of characters of each picture through segmentation, performing direction recognition on a detected text region, and screening texts in different directions;
step 1.3: and for the voltage level of the graphic element, the crossing relation and the rotation angle of the bus, forming a three-element candidate topology grouping of the main wiring diagram based on the topology relation obtained from the identification result and the image processing, carrying out text and graphic element association relation coverage relation inspection and topology connectivity judgment, merging and dividing the topology grouping, and determining the voltage level of graphic element equipment and the associated text attribute of the graphic element equipment in the topology grouping.
In the step 2, comparing and analyzing the main wiring diagram recognition result with the CAD graphic file, visualizing the difference element, updating the main wiring diagram recognition result according to the revision information input by the user, and further comprising:
performing region contrast analysis on the identified graphic element, connection line, text element and the content of the corresponding position in the original image based on the image similarity, and returning the difference part as candidate difference element and analysis result data to the front-end tool;
taking the original CAD graph file as a base map, superposing and drawing an analysis result, visualizing the analysis result, and highlighting candidate difference elements; and editing the elements by combining the revision information, correcting the conditions of missing marks, wrong marks and multiple marks, and updating the identification result.
The step 4 further comprises:
transmitting the power graphic file of the transformer substation monitoring system to a front-end tool through calling parameters, and generating model data according to the updated graphic file of the current transformer substation monitoring system;
the model data and the current model data of the system are transmitted to the algorithm service back end for comparison and verification to obtain the difference part of the model data and the current model data,
and returning the difference part to the front-end tool for visual display, highlighting the difference part of the two models, receiving confirmation of the user on the difference part, and forming a verification report and export.
The invention also discloses a pattern generating device based on the main wiring diagram recognition technology by utilizing the pattern generating method based on the main wiring diagram recognition technology, which comprises a main wiring diagram element recognition module, a recognition result revision conversion module, a recognition model lifting module and a pattern checking module.
The main wiring diagram element identification module is used for transmitting a CAD graphic file to be analyzed in the transformer substation monitoring system to the algorithm service rear end through the front end tool, and identifying main wiring diagram elements of the CAD graphic file based on an identification model of the algorithm service rear end to obtain primitive information, text information and topology information;
the identification result revision conversion module is used for comparing and analyzing the identification result of the main wiring diagram with the CAD graphic file, visualizing the difference elements, updating the identification result of the main wiring diagram according to revision information input by a user, converting the identification result of the main wiring diagram into a power graphic file and a power model file, and returning to the substation monitoring system;
the identification model lifting module is used for feeding back the revision information to the algorithm service back end and updating and training the identification model;
and the graph model checking module is used for regenerating the power model data according to the power pattern file stored by the transformer substation monitoring system, comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end, and processing model differences.
Correspondingly, the application also discloses a terminal, which comprises a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is operative in accordance with the instructions to perform the steps of the pattern generation method in accordance with the primary wiring diagram identification technique described above.
Accordingly, the present application also discloses a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the aforementioned pattern generation method based on the primary wiring pattern recognition technique.
Compared with the prior art, the invention provides the pattern generation method and the device based on the main wiring diagram recognition technology, which utilize the artificial intelligence technology to perform arrangement recognition of elements such as equipment, connection lines, characters and the like on CAD pictures which are referenced when the scheduling technical support system is drawn, automatically generate electric power pattern data and electric power pattern data, lighten repeated work of automation personnel and realize the penetration from static pictures to online monitoring pictures. Meanwhile, by utilizing a mechanism combining automatic identification and manual checking and a processing mechanism for identifying errors, the system identification result can be ensured to be completely correct, the system can automatically iterate, and the identification accuracy is improved.
Drawings
Fig. 1 is a schematic diagram of a pattern generation system based on a primary wiring diagram recognition technique in the present invention.
Fig. 2 is a flowchart of a pattern generation method based on the primary wiring diagram recognition technology in the present invention.
Fig. 3 is a collaborative flow diagram of pattern generation based on primary wiring diagram identification techniques in the present invention.
Fig. 4 is a collaborative flow chart of pattern check based on primary wiring pattern recognition techniques in the present invention.
Fig. 5 is a schematic structural diagram of a pattern generating device based on the primary wiring pattern recognition technology in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The embodiments described herein are merely some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art without making any inventive effort, are within the scope of the present invention.
Aiming at the defects of the prior art, the invention provides a pattern generation method and a pattern generation device based on a main wiring pattern recognition technology, which utilize an artificial intelligence technology to perform arrangement recognition of elements such as equipment, connection lines, characters and the like on CAD pictures which are referred to when a dispatching technical support system is drawn, and automatically generate electric power pattern data and electric power model data.
As shown in fig. 1, the pattern generation system based on the primary wiring pattern recognition technology includes two parts: the front-end tool and the algorithm service back-end cooperate with the substation monitoring system to complete the graph model generating task. The algorithm service back end operates on an independent algorithm server, provides main wiring diagram identification and topology calculation services, and has a model lifting training function. The front-end tool is deployed in the substation monitoring system and used for executing functions of CAD graph and stock substation graph import, recognition result visualization, graph inspection, annotation result file export and the like. The substation monitoring system and the front-end tool communicate through process calling, and the graphic model file is transmitted to the front-end tool through calling parameters to execute the graphic model generation flow. And the front-end tool acquires the CAD graph, and transmits the analysis task and the analysis data to the algorithm service back end in a message queue mode for asynchronous analysis and calling. After the analysis result is received, the user performs manual check, and the graphic model file is returned to the substation monitoring system so as to perform graphic model checking and warehousing.
Referring to fig. 2, the pattern generation method based on the main wiring diagram recognition technology disclosed by the invention comprises the following steps:
step 1: and transmitting the CAD graphic file to be analyzed in the substation monitoring system to an algorithm service rear end through a front end tool, and identifying the main wiring diagram elements of the CAD graphic file based on an identification model of the algorithm service rear end to obtain graphic element information, text information and topology information.
In the CAD graph recognition scene, the substation monitoring system transmits the designed CAD graph file to the algorithm service rear end through the front end tool, and the rear end starts main element analysis of the main wiring graph, including graphic element analysis, text analysis and topology analysis, and is used for equipment, connection lines, characters and the like in the main wiring graph.
The step 1 further comprises:
step 1.1: and carrying out primitive identification in primitive analysis, and carrying out deformation and enhancement treatment on limited training samples so as to form an identification model with stronger generalization capability. And realizing the ultra-high resolution primitive identification by adopting a fusion algorithm based on an image segmentation pyramid.
The original is first divided into a plurality of n×n sub-graphs, where N is a prime number, n= [2,3,5, 7..max N ]，Max N Ceil { min (img_w, img_h)/320 }, where img_w, img_h represent artwork resolution. During the division, a certain overlap region remains between the sub-images, which is a settable value (for example, the lateral overlap region is set to 5% of the width of the sub-images).
Inputting all the subgraphs into a primitive identification model for identification, obtaining the types of the primitives and the coordinates of bounding boxes of the primitives after identification, and restoring the coordinates of the subgraphs after segmentation of a certain layer to the coordinates of the original drawings, namely restoring the coordinates of the bounding boxes in the identification result of the segmented subgraphs of a certain layer to the coordinates of the whole original drawings, and further executing the combination and filtering of the results.
The merging process comprises merging segmentation recognition results of the same primitive aiming at the edge region of the subgraph, wherein the precondition is that the categories are the same, the bounding boxes are the same in width or height and are closely attached; the filtering process includes, for the same primitive, possibly having repeated recognition results on multiple layers of the image pyramid, filtering the repeated recognition results, where the filtering rule is multiple detection results with the same category of detection results and the IOU meeting a certain set threshold, where: IOU = the area of the intersection part of two bounding boxes/the area of the union part of two bounding boxes.
And selecting the detection result with the highest confidence coefficient, and deleting other detection results to obtain the final primitive detection result.
Step 1.2: character recognition is carried out in text analysis, and aiming at the situation that some text fonts in a main wiring diagram are special, transfer learning and customized training are carried out on the basis of the existing OCR recognition model, and meanwhile, mutual verification and correlation are carried out by combining topological relations and primitive elements. And performing mask operation on all the detected primitives, namely filling the region where the primitives are positioned with pure colors. The number of characters per figure is reduced by segmentation by an image pyramid method similar to that in step 1.1. In addition, the original picture is rotated by 90 degrees and positive and negative, so that vertical characters rotating by positive 90 degrees and negative 90 degrees can be conveniently identified. And after the detection stage is finished, adding direction recognition of the detected text region for screening texts in different directions.
Step 1.3: in image processing and topology analysis, three-element candidate topology groupings of a main wiring diagram are formed based on the topology relationship obtained from the identification result and the image processing for the voltage level, the crossing relationship and the rotation angle of the buses and the like of the graphic elements, and the association relationship and the coverage relationship of texts and the graphic elements are checked. The checking of the association relation refers to checking of naming rules of texts associated with the primitive types. For example, "2220" is the text number of a certain breaker, and the device number of the breaker with the number "2220" in one topology group should be "-17", "-57", "-5", "-4" or "2220-17", "2220-57", "2220-5", "2220-4" format, and the text associated with the analyzed primitive needs to be checked according to this naming rule. Checking the coverage relation refers to checking whether the graphic elements and the texts are in one-to-one correspondence, and if the graphic elements or the texts which are not in correspondence exist, warning prompt is needed.
And identifying the graphic element, the text and the connecting wire as independent main wiring graphic elements, and associating the graphic element with the text through association of the graphic element with the text. And (3) starting to combine topology groups from independent connecting wires, firstly combining connecting wires with adjacent head and tail ends of the connecting wires into respective candidate topology groups, and further combining pictures with adjacent end points of any connecting wire in the group into the groups. Then, because the main wiring diagram has the scene that the connection lines between the graphic elements cross the bus, the connection lines crossing the bus need to be analyzed according to the topological diagram structure, and the specific steps are to carry out recursive search on the two ends of the connection lines crossing the bus, when the graphic elements connected with the two ends of the line have no grounding disconnecting link or line terminal, the connection lines are judged to be actually crossed but not crossed with the bus, the connection lines need to be split, namely the topology groups of the connection lines need to be split into two topology groups.
And determining the voltage level of the graphic element equipment in the topology group according to the finally formed topology group. When determining the voltage class of each graphic element, starting from a bus or line terminal containing a voltage class attribute, performing voltage class assignment on the graphic elements in a connection group connecting the buses, and performing voltage class amplitude recursively on other graphic elements in the group where the graphic elements are located. The recursion is stopped if a transformer is encountered. Because the transformer will connect topology groupings of multiple voltage classes at the same time, the voltage class of the transformer is determined to be the maximum of the connected topology groupings.
In the topology connectivity judging step, firstly, one topology group is arbitrarily selected for recursive traversal: traversing each primitive of the group in turn, and when the primitives belong to other topological groups at the same time, traversing the other topological groups until all the associated topological groups are traversed. If there is no topology grouping traversed at this time, it is indicated that there is topology analysis that does not establish a connection with the master wiring diagram, and an alarm prompt needs to be given.
Step 2: and comparing and analyzing the main wiring diagram identification result with the CAD graph file, visualizing the difference elements, updating the main wiring diagram identification result according to revision information input by a user, converting the main wiring diagram identification result into a power graph file and a power model file, and returning to the substation monitoring system.
Fig. 3 is a collaborative flow diagram of pattern generation based on primary wiring diagram identification techniques in the present invention. According to a further preferred embodiment, after generating the analysis result data, entering an image element auxiliary self-checking process, performing area contrast analysis on the identified graphic element, connection line, text element and the content of the corresponding position in the original image based on the image similarity, and returning the difference part as a candidate difference element and the analysis result data to the front end tool. For example, for a certain primitive detection result, selecting a template primitive consistent with the detection result type from a standard primitive library, intercepting the primitive in the original primitive according to the coordinates of a bounding box, respectively performing image binarization on the intercepted primitive subgraph and the template primitive, performing pixel level comparison, judging that the current prediction result is likely to have errors if different pixel duty ratios reach a certain threshold value, marking the current prediction result as a candidate difference element, and returning the candidate difference element to a front-end tool for manually performing final judgment.
The front-end tool visualizes the analysis result data and the candidate discrepancy element. Specifically, the original CAD graph file is taken as a base map, the analysis results are overlapped and drawn, the visualization of the analysis results is carried out, and the warning area of the self-checking candidate difference element is highlighted. And combining the self-checking result to artificially check the correctness of the original recognition result, editing the problematic elements, correcting the conditions of missing marks, wrong marks and multiple marks, and further updating the recognition result.
Meanwhile, the front-end tool records and stores the error elements as error correction samples to be input into a sample management database in the model training module for subsequent lifting training of the model.
After the manual checking and revising of the front-end tool, the intermediate recognition result is converted into a standard power graph file and a standard power model file according to the power graph model data format, and the standard power graph file and the standard power model file are returned to the substation monitoring system.
Step 3: and feeding the revision information back to the algorithm service back end, and updating and training the identification model.
The front end tool can be used for optimizing and improving the back end recognition model while being manually revised and returned to the monitoring system, and the method specifically comprises the following steps:
in the manual revision process of pattern recognition, elements with wrong recognition (including missed marks and wrong marks) are extracted and transmitted to the algorithm service back end. And the algorithm service back end extracts the received data target and generates samples, and supplements the received data target and the samples into a training sample library. The model training module carries out incremental training on the identification model according to a determined strategy (a certain number of instant incremental samples is reached), and the model in the model library is updated after training is completed.
Step 4: and regenerating power model data according to the power pattern file stored by the transformer substation monitoring system, and comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end to process model differences.
Fig. 4 is a collaborative flow chart of pattern check based on primary wiring pattern recognition techniques in the present invention. After the substation monitoring system receives the identified power graphic file and the power model file, a diagram model checking process can be executed by combining the inventory substation graphic file, and the method specifically comprises the following steps:
the method comprises the steps that a graph model checking module of a transformer substation monitoring system transmits a power graph file to a front-end tool through calling parameters, model data are generated according to the graph file updated by the current transformer substation monitoring system, the model data and the current model data of the system are transmitted to an algorithm service rear end for comparison and checking, difference parts of the model data and the model data are obtained, the difference parts are returned to the front-end tool for visual display, and the difference parts of the two models are highlighted. And receiving confirmation of the user on the difference part to form a verification report export.
Compared with the prior art, the invention provides the pattern generation method and the device based on the main wiring diagram recognition technology, which utilize the artificial intelligence technology to perform arrangement recognition of elements such as equipment, connection lines, characters and the like on CAD pictures which are referenced when the scheduling technical support system is drawn, automatically generate electric power pattern data and electric power pattern data, lighten repeated work of automation personnel and realize the penetration from static pictures to online monitoring pictures. Meanwhile, by utilizing a mechanism combining automatic identification and manual checking and a processing mechanism for identifying errors, the system identification result can be ensured to be completely correct, the system can automatically iterate, and the identification accuracy is improved.
The present invention may be a system, method, and/or computer program product. Referring to fig. 5, the invention also discloses a pattern generation device based on the main wiring diagram recognition technology based on the pattern generation method based on the main wiring diagram recognition technology, which comprises a main wiring diagram element recognition module 1, a recognition result revision conversion module 2, a recognition model lifting module 3 and a pattern checking module 4.
The main wiring diagram element identification module 1 is used for transmitting a CAD graphic file to be analyzed in the transformer substation monitoring system to the algorithm service rear end through a front end tool, and identifying main wiring diagram elements of the CAD graphic file based on an identification model of the algorithm service rear end to obtain primitive information, text information and topology information;
the recognition result revision conversion module 2 is used for comparing and analyzing the recognition result of the main wiring diagram with the CAD graphic file, visualizing the difference elements, updating the recognition result of the main wiring diagram according to revision information input by a user, converting the recognition result of the main wiring diagram into a power graphic file and a power model file, and returning to the substation monitoring system;
the recognition model lifting module 3 is used for feeding back the revision information to the algorithm service back end and updating and training the recognition model;
and the graph model checking module 4 is used for regenerating the power model data according to the power graph file stored by the transformer substation monitoring system, comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end, and processing model differences.
Based on the spirit of the present invention, one skilled in the art can easily appreciate that a computer program product can be obtained based on the aforementioned pattern generation method based on the primary wiring pattern recognition technique. The computer program product may include a computer readable storage medium having computer readable program instructions embodied thereon for causing a processor to implement aspects of the present disclosure. The application also comprises a terminal, which comprises a processor and a storage medium; the storage medium is used for storing instructions; the processor is operative in accordance with the instructions to perform the steps of the pattern generation method in accordance with the primary wiring diagram identification technique described above.
The computer readable storage medium may be a tangible device that can hold and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: portable computer disks, hard disks, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static Random Access Memory (SRAM), portable compact disk read-only memory (CD-ROM), digital Versatile Disks (DVD), memory sticks, floppy disks, mechanical coding devices, punch cards or in-groove structures such as punch cards or grooves having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media, as used herein, are not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., optical pulses through fiber optic cables), or electrical signals transmitted through wires.
The computer readable program instructions described herein may be downloaded from a computer readable storage medium to a respective computing/processing device or to an external computer or external storage device over a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers and/or edge servers. The network interface card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium in the respective computing/processing device.
Computer program instructions for performing the operations of the present disclosure can be assembly instructions, instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, c++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer readable program instructions may be executed entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of remote computers, the remote computer may be connected to the user computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (e.g., connected through the internet using an internet service provider). In some embodiments, aspects of the present disclosure are implemented by personalizing electronic circuitry, such as programmable logic circuitry, field Programmable Gate Arrays (FPGAs), or Programmable Logic Arrays (PLAs), with state information of computer readable program instructions, which can execute the computer readable program instructions.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that: modifications and equivalents may be made to the specific embodiments of the invention without departing from the spirit and scope of the invention, which is intended to be covered by the claims.
Claims (12)
1. The pattern generation method based on the main wiring diagram recognition technology is characterized by comprising the following steps of:
step 1, a CAD graph file to be analyzed in a transformer substation monitoring system is transmitted to an algorithm service rear end through a front end tool, and main wiring graph elements of the CAD graph file are identified based on an identification model of the algorithm service rear end, so that graphic element information, text information and topology information are obtained;
step 2, comparing and analyzing the main wiring diagram recognition result with the CAD graph file, visualizing the difference elements, updating the main wiring diagram recognition result according to revision information input by a user, converting the main wiring diagram recognition result into a power graph file and a power model file, and returning to the substation monitoring system;
step 3, feeding back the revised information to the algorithm service back end, and updating and training the identification model;
and 4, regenerating power model data according to the power pattern file stored by the transformer substation monitoring system, and comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end to process model differences.
2. The method for generating a pattern based on the primary wiring diagram recognition technology according to claim 1, wherein in the step 1, a CAD pattern file to be analyzed in the substation monitoring system is transferred to the algorithm service back end through a front end tool, and further comprising:
the front-end tool is deployed in the transformer substation monitoring system, the transformer substation monitoring system and the front-end tool communicate through process calling, files are transmitted to the front-end tool through calling parameters, the front-end tool acquires CAD graphic files, analysis tasks and analysis data are transmitted to the algorithm service rear end through a message queue mode to carry out asynchronous analysis calling, and the algorithm service rear end operates on an independent algorithm server.
3. The pattern generation method based on the primary wiring diagram recognition technology according to claim 2, wherein in the step 1, primary wiring diagram elements of the CAD graphic file are recognized to obtain primitive information, text information and topology information, and further comprising:
step 1.1: carrying out deformation and enhancement treatment on the training sample, adopting an identification algorithm based on an image segmentation pyramid to realize the identification of the ultra-high resolution main wiring diagram, fusing sub-graph identification results under each scale, and further screening and de-duplicating the multi-scale identification results to obtain a final identification result;
step 1.2: performing transfer learning and customized training on the OCR recognition model, performing mutual verification and association by combining a topological relation and primitive elements, deleting pictures of equipment primitives, reducing the number of characters of each picture through segmentation, performing direction recognition on a detected text region, and screening texts in different directions;
step 1.3: and for the voltage level of the graphic element, the crossing relation and the rotation angle of the bus, forming a three-element candidate topology grouping of the main wiring diagram based on the topology relation obtained from the identification result and the image processing, carrying out text and graphic element association relation coverage relation inspection and topology connectivity judgment, merging and dividing the topology grouping, and determining the voltage level of graphic element equipment and the associated text attribute of the graphic element equipment in the topology grouping.
4. A pattern generation method based on a primary wiring diagram recognition technology according to claim 3, wherein in the step 2, a primary wiring diagram recognition result is compared with the CAD pattern file, a difference element is visualized, and the primary wiring diagram recognition result is updated according to revision information input by a user, and further comprising:
performing region contrast analysis on the identified graphic element, connection line, text element and the content of the corresponding position in the original image based on the image similarity, and returning the difference part as candidate difference element and analysis result data to the front-end tool;
taking the original CAD graph file as a base map, superposing and drawing an analysis result, visualizing the analysis result, and highlighting candidate difference elements; and editing the elements by combining the revision information, correcting the conditions of missing marks, wrong marks and multiple marks, and updating the identification result.
5. The pattern generation method based on the main wiring diagram recognition technology according to claim 4, wherein the step 4 further comprises:
transmitting the power graphic file of the transformer substation monitoring system to a front-end tool through calling parameters, and generating model data according to the updated graphic file of the current transformer substation monitoring system;
the model data and the current model data of the system are transmitted to the algorithm service back end for comparison and verification to obtain the difference part of the model data and the current model data,
and returning the difference part to the front-end tool for visual display, highlighting the difference part of the two models, receiving confirmation of the user on the difference part, and forming a verification report and export.
6. A pattern generation device based on a primary wiring diagram recognition technology by using the pattern generation method based on a primary wiring diagram recognition technology according to any one of claims 1 to 5, comprising a primary wiring diagram element recognition module, a recognition result revision conversion module, a recognition model promotion module, and a pattern check module, wherein:
the main wiring diagram element identification module is used for transmitting a CAD graphic file to be analyzed in the transformer substation monitoring system to the algorithm service rear end through the front end tool, and identifying main wiring diagram elements of the CAD graphic file based on an identification model of the algorithm service rear end to obtain primitive information, text information and topology information;
the identification result revision conversion module is used for comparing and analyzing the identification result of the main wiring diagram with the CAD graphic file, visualizing the difference elements, updating the identification result of the main wiring diagram according to revision information input by a user, converting the identification result of the main wiring diagram into a power graphic file and a power model file, and returning to the substation monitoring system;
the identification model lifting module is used for feeding back the revision information to the algorithm service back end and updating and training the identification model;
and the graph model checking module is used for regenerating the power model data according to the power pattern file stored by the transformer substation monitoring system, comparing and checking the regenerated power model data with the power model file by utilizing the algorithm service back end, and processing model differences.
7. The pattern generation device based on the primary wiring diagram recognition technology according to claim 6, wherein the front-end tool is deployed in the substation monitoring system, the substation monitoring system and the front-end tool communicate through process call, files are transmitted to the front-end tool through call parameters, the front-end tool acquires CAD graphic files, analysis tasks and analysis data are transmitted to an algorithm service back-end through a message queue mode for asynchronous analysis call, and the algorithm service back-end operates on an independent algorithm server.
8. The pattern generation apparatus based on a primary wiring diagram recognition technique of claim 7, wherein the primary wiring diagram element recognition module is further configured to:
carrying out deformation and enhancement treatment on the training sample, adopting an identification algorithm based on an image segmentation pyramid to realize the identification of the ultra-high resolution main wiring diagram, fusing sub-graph identification results under each scale, and further screening and de-duplicating the multi-scale identification results to obtain a final identification result;
performing transfer learning and customized training on the OCR recognition model, performing mutual verification and association by combining a topological relation and primitive elements, deleting pictures of equipment primitives, reducing the number of characters of each picture through segmentation, performing direction recognition on a detected text region, and screening texts in different directions;
and for the voltage level of the graphic element, the crossing relation and the rotation angle of the bus, forming a three-element candidate topology grouping of the main wiring diagram based on the topology relation obtained from the identification result and the image processing, carrying out text and graphic element association relation coverage relation inspection and topology connectivity judgment, merging and dividing the topology grouping, and determining the voltage level of graphic element equipment and the associated text attribute of the graphic element equipment in the topology grouping.
9. The pattern generation device based on the main wiring diagram recognition technology according to claim 8, wherein the recognition result revision conversion module is further configured to:
performing region contrast analysis on the identified graphic element, connection line, text element and the content of the corresponding position in the original image based on the image similarity, and returning the difference part as candidate difference element and analysis result data to the front-end tool;
taking the original CAD graph file as a base map, superposing and drawing an analysis result, visualizing the analysis result, and highlighting candidate difference elements; and editing the elements by combining the revision information, correcting the conditions of missing marks, wrong marks and multiple marks, and updating the identification result.
10. The pattern generation device based on the primary wiring diagram recognition technology according to claim 9, wherein the pattern check module is further configured to:
performing region contrast analysis on the identified graphic element, connection line, text element and the content of the corresponding position in the original image based on the image similarity, and returning the difference part as candidate difference element and analysis result data to the front-end tool;
taking the original CAD graph file as a base map, superposing and drawing an analysis result, visualizing the analysis result, and highlighting candidate difference elements; and editing the elements by combining the revision information, correcting the conditions of missing marks, wrong marks and multiple marks, and updating the identification result.
11. A terminal comprising a processor and a storage medium; the method is characterized in that:
the storage medium is used for storing instructions;
the processor is operative according to the instructions to perform the steps of the pattern generation method based on the primary wiring diagram identification technique according to any one of claims 1-5.
12. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the pattern generation method based on the primary wiring pattern recognition technique as claimed in any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311445433.3A CN117612195A (en) | 2023-11-01 | 2023-11-01 | Graph model generation method and device based on main wiring graph recognition technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311445433.3A CN117612195A (en) | 2023-11-01 | 2023-11-01 | Graph model generation method and device based on main wiring graph recognition technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117612195A true CN117612195A (en) | 2024-02-27 |
Family
ID=89958666
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311445433.3A Pending CN117612195A (en) | 2023-11-01 | 2023-11-01 | Graph model generation method and device based on main wiring graph recognition technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117612195A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118038481A (en) * | 2024-04-12 | 2024-05-14 | 江苏省送变电有限公司 | A method and system for identifying substation terminal block drawings based on computer vision technology |
-
2023
- 2023-11-01 CN CN202311445433.3A patent/CN117612195A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118038481A (en) * | 2024-04-12 | 2024-05-14 | 江苏省送变电有限公司 | A method and system for identifying substation terminal block drawings based on computer vision technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113255499B (en) | Digital automatic modeling method for secondary loop of transformer substation cable | |
| CN113485160A (en) | Simulation modeling method and device based on pattern matching recognition | |
| US20250029367A1 (en) | Component identification method and apparatus, electronic device, and storage medium | |
| CN116541912B (en) | Method and device for generating wiring diagram based on CAD image recognition | |
| EP3933653A1 (en) | Method and system for generating a digital representation of one or more engineering diagrams in a computing environment | |
| CN120450895A (en) | A system and method for automatic analysis of power grid wiring diagrams based on deep learning | |
| CN117608545B (en) | Standard operation program generation method based on knowledge graph | |
| CN115861609A (en) | Segmentation and labeling method, electronic equipment and storage medium of remote sensing image | |
| CN115880103A (en) | Visual management method, device, equipment and medium for railway engineering progress | |
| CN117612195A (en) | Graph model generation method and device based on main wiring graph recognition technology | |
| CN116052193B (en) | Method and system for picking and matching dynamic tables in RPA interface | |
| CN117671718A (en) | PLC wiring diagram automatic identification and physical information extraction method, device and equipment | |
| CN117078184A (en) | Drawing examination method, system and terminal based on artificial intelligence algorithm | |
| CN119964189B (en) | Automatic identifying method, system, medium and equipment for station wiring diagram link relation | |
| Putra et al. | Automatic extraction of cable connection information from 2D drawings for electrical outfittings design in shipyards | |
| US20240344832A1 (en) | Training method for map-generation large model and map generation method | |
| CN115984838A (en) | POI name generation method, device, electronic device and storage medium | |
| JP3078038B2 (en) | Automatic drawing input system and method | |
| CN120296082B (en) | Signal flow graph visualization method and system for nuclear power plants | |
| CN121327156B (en) | A method and system for intelligent government writing based on multi-agent collaboration and long memory technology | |
| CN120632971B (en) | Intelligent power distribution network mapping method, system and medium based on SIFT algorithm | |
| CN111462107A (en) | End-to-end high-precision industrial part shape modeling method | |
| CN116740112B (en) | Numbering method, positioning method, device, equipment and medium for UI (user interface) element | |
| CN117372510B (en) | Map annotation identification method, terminal and medium based on computer vision model | |
| CN115565201B (en) | Taboo picture identification method, apparatus and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |