CN117975729A

CN117975729A - Traffic information issuing system, method, computer device and storage medium

Info

Publication number: CN117975729A
Application number: CN202410243574.5A
Authority: CN
Inventors: 朱艳玲; 师坤
Original assignee: Shandong Ouhua Information Technology Co ltd
Current assignee: Shandong Ouhua Information Technology Co ltd
Priority date: 2024-03-04
Filing date: 2024-03-04
Publication date: 2024-05-03
Anticipated expiration: 2044-03-04
Also published as: CN117975729B

Abstract

The invention relates to the technical field of intelligent transportation, in particular to a traffic information release system, a method, computer equipment and a storage medium. In the invention, the traffic information priority is accurately assessed through a multi-attribute decision making technology, the emergency information is ensured to be issued preferentially, the real-time performance and the accuracy of traffic management are improved, the user behavior is analyzed by adopting an isolated forest and a clustering algorithm, the response capability and the prediction accuracy to an emergency event are improved, the traffic flow and the congestion situation are comprehensively predicted by combining an autoregressive moving average model and a circulating neural network, the potential congestion point can be predicted and evaluated by a graph convolution network, scientific basis is provided for traffic management, the traffic information recommendation and the optimized issuing strategy are personalized, and the user experience and the information service efficiency are improved.

Description

Traffic information issuing system, method, computer device and storage medium

Technical Field

The present invention relates to the field of intelligent traffic technologies, and in particular, to a traffic information issuing system, method, computer device, and storage medium.

Background

The intelligent transportation is a comprehensive transportation management system which applies modern information technology, data communication transmission technology, electronic perception technology, control technology and computer technology to the whole ground traffic management system, and through comprehensive application of the technologies, the traffic efficiency is improved, the traffic safety is enhanced, the traffic service is improved, the traffic time is saved, and the environmental impact is reduced.

The system is designed to improve traffic efficiency, reduce traffic jams, improve road safety, and provide support for users to plan optimal routes, thereby achieving the effects of relieving traffic pressure, improving public traffic utilization rate, reducing accident occurrence rate, and the like.

The traditional traffic information release system lacks deep analysis of user behaviors and comprehensive integration of traffic data, so that limitation exists in traffic management and information release, traffic flow and congestion points cannot be accurately predicted, effective measures cannot be taken in advance, traffic congestion is relieved or avoided, traffic efficiency and user travel experience are affected, personalized traffic information recommendation is lacking, users cannot obtain travel suggestions which are most suitable for the users, travel efficiency is low, the traditional system generally adopts a static mode in an information release strategy, capability of adjusting according to real-time traffic conditions and user requirements is lacking, practicality and effectiveness of information service are reduced, user experience is poor, and overall quality of public service is reduced.

Disclosure of Invention

The invention aims to solve the defects in the prior art, and provides a traffic information issuing system, a method, computer equipment and a storage medium.

In order to achieve the above purpose, the present invention adopts the following technical scheme: the traffic information issuing system comprises an information evaluation module, a user behavior analysis module, a cross-domain data analysis module, a correlation analysis module, an information recommendation module and an issuing strategy optimization module;

the information evaluation module quantitatively scores the urgency degree, the influence range and the time sensitivity attribute of the information by adopting a multi-attribute decision making technology based on the collected traffic information, performs weight distribution and comprehensive scoring on the multi-item attribute by adopting a weighting and sequencing method, sequences each piece of information, optimizes the sequencing result by adopting a technical sequencing method, determines the information priority, converts the multi-dimensional information attribute into a single comprehensive score and generates an information priority evaluation result;

The user behavior analysis module analyzes user behavior data by adopting an isolated forest algorithm based on an information priority evaluation result, identifies abnormal behaviors, subdivides the abnormal behavior patterns by a clustering algorithm, identifies atypical query patterns and demand mutation, refines information, and generates a user behavior abnormal analysis result;

The cross-domain data analysis module processes historical traffic data by adopting an autoregressive moving average model based on a user behavior anomaly analysis result, analyzes trend and seasonal change, predicts traffic flow, integrates traffic monitoring, weather and social activity cross-domain data by adopting a cyclic neural network technology, predicts traffic flow and congestion conditions by deep learning, analyzes and extracts cross-domain data correlation, and generates a traffic prediction analysis scheme;

The association analysis module is used for analyzing the association between a plurality of nodes and paths in the traffic network through a graph rolling network based on a traffic prediction analysis scheme, analyzing the dependency relationship of the nodes, predicting potential congestion points and influence on surrounding traffic flows, revealing the distribution characteristics and dynamic change rules of the traffic flows and generating a traffic flow association analysis result;

The information recommendation module analyzes the travel habit and preference of the user by adopting a convolutional neural network based on the traffic flow relevance analysis result and the user behavior abnormality analysis result, learns the travel mode, processes real-time traffic condition data by utilizing the convolutional neural network, predicts traffic condition change in a short period, analyzes the user requirements and the current traffic condition and generates a personalized traffic information recommendation result;

The issuing strategy optimizing module optimizes the content, frequency and channel of information issuing through a dynamic adjusting algorithm based on personalized traffic information recommending results, analyzes user feedback and real-time traffic condition change, adjusts the issuing strategy to match current requirements, optimizes timeliness and relativity of information issuing, and generates an information issuing strategy;

The information priority evaluation result comprises an emergency degree index, an influence area size and an expected influence duration, the user behavior abnormality analysis result comprises an abnormality query frequency, an abnormality query time period and atypical route selection, the traffic prediction analysis scheme comprises predicted traffic flow change, potential congestion areas and key external factors influencing traffic flow, the traffic flow relevance analysis result comprises relevance scores of key nodes and paths, predicted congestion nodes and traffic flow transmission paths, the personalized traffic information recommendation result comprises a user preference route, a recommended travel time period and a coping strategy, and the information release strategy comprises information content adjustment parameters, adjustment rules of release frequencies and selected release channels.

As a further scheme of the invention, the information evaluation module comprises an emergency degree analysis sub-module, an influence range analysis sub-module and a time sensitivity analysis sub-module;

The emergency degree analysis submodule quantitatively scores the emergency degree of the information by adopting a multi-attribute decision making technology based on the collected traffic information, converts the emergency condition of the information into a numerical score by setting an emergency degree quantization standard, and generates an emergency degree score;

The influence range analysis submodule quantitatively scores the influence range of the information by adopting a weighting and sorting method based on the emergency degree score, and performs weight distribution by the emergency degree score to generate an influence range score;

The time sensitivity analysis submodule adopts a technical sequencing method to quantitatively score the time sensitivity of the information based on the influence range score, comprehensively analyzes the emergency degree score and the influence range score to determine the priority of the information and generate an information priority assessment result.

As a further scheme of the invention, the user behavior analysis module comprises a query frequency analysis sub-module, a query time analysis sub-module and a route selection analysis sub-module;

The inquiry frequency analysis submodule analyzes inquiry frequencies of users by adopting an isolated forest algorithm based on the information priority evaluation result, screens inquiry behaviors of abnormal frequencies and generates inquiry frequency abnormal indexes;

The inquiry time analysis submodule analyzes the time distribution of the inquiry behaviors of the user by adopting a clustering algorithm based on the inquiry frequency abnormal indexes, identifies the abnormal inquiry behaviors in atypical time periods, identifies the effect of time factors on the abnormal behaviors of the user and generates inquiry time abnormal indexes;

The route selection analysis submodule analyzes the route selection behavior of the user by adopting an isolated forest algorithm based on the abnormal index of the query time, identifies the atypical query mode of the user, evaluates the abnormal degree of the user behavior by the query frequency of the user and the abnormal index of the query time and generates an abnormal analysis result of the user behavior.

As a further scheme of the invention, the cross-domain data analysis module comprises a traffic flow prediction sub-module, a congestion point identification result sub-module and an external factor influence analysis sub-module;

The traffic flow prediction submodule adopts an autoregressive moving average model based on the analysis result of the abnormal behavior of the user, analyzes the trend and seasonal change of the traffic flow through historical traffic data, establishes a model through the historical data, predicts the future traffic flow and generates traffic flow trend prediction;

The congestion point identification result submodule is used for integrating real-time traffic monitoring data based on traffic flow trend prediction by adopting a cyclic neural network technology, analyzing and predicting congestion points, processing time sequence data, identifying areas and time of flow sudden increase and generating a congestion point identification result;

the external factor influence analysis submodule integrates cross-domain data based on the congestion point identification result by adopting a cyclic neural network technology, analyzes weather and social activities affecting traffic flow and congestion conditions, predicts the influence of the cross-domain data on the traffic flow by a deep learning model, and generates a traffic prediction analysis scheme.

As a further scheme of the invention, the association analysis module comprises a node association evaluation sub-module, a congestion prediction sub-module and a traffic propagation analysis sub-module;

The node relevance scoring submodule analyzes relevance among a plurality of nodes in a traffic network by adopting a graph rolling network based on a traffic prediction analysis scheme, evaluates importance and relevance of the nodes in the traffic network, and generates node relevance scores by extracting node characteristics, aggregating neighbor information and calculating relevance scores;

The congestion prediction submodule adopts a long-period memory network to evaluate potential congestion points based on node relevance scores, captures long-period dependency relations in time sequence data, analyzes traffic flow data and node relevance, predicts congestion conditions occurring in a target time period and generates congestion point prediction information;

the traffic flow propagation analysis submodule analyzes the propagation modes of traffic flow, including the relevance among nodes and traffic flow change, based on the congestion point prediction information, applies an infectious disease propagation model, analyzes the distribution characteristics and dynamic change rules of traffic flow in the whole network and generates a traffic flow relevance analysis result.

As a further scheme of the invention, the information recommendation module comprises a travel habit analysis sub-module, a real-time condition analysis sub-module and a recommendation strategy generation sub-module;

the travel habit analysis submodule analyzes user history travel data based on a traffic flow relevance analysis result and a user behavior abnormality analysis result by adopting a convolutional neural network technology, identifies user travel habits and preferences, and generates a user travel habit analysis result through a user history travel mode comprising a common route and travel time;

the real-time condition analysis submodule analyzes the current traffic condition, including congestion information, accident information and user travel habits, by adopting a circulating neural network technology and through real-time traffic condition data based on the user travel habit analysis result to generate a real-time traffic condition analysis result;

The recommendation strategy generation submodule is used for analyzing travel time and route through user travel habits and current traffic conditions by adopting a convolutional neural network and a cyclic neural network technology based on real-time traffic condition analysis results to generate personalized traffic information recommendation results.

As a further scheme of the invention, the distribution strategy optimization module comprises a content adjustment strategy sub-module, a distribution frequency management sub-module and a channel selection optimization sub-module;

The content adjustment strategy sub-module extracts user characteristics by adopting a text mining algorithm based on personalized traffic information recommendation results, classifies and prioritizes information by using a support vector machine algorithm through historical preference and feedback of a user, analyzes information content, performs personalized adjustment and generates a content adjustment scheme;

The release frequency management submodule is based on a content adjustment scheme, adopts time sequence analysis, matches information release time points, adjusts information release frequency, adjusts release frequency through a poisson distribution model according to a prediction result and real-time traffic conditions, and generates a frequency adjustment scheme;

The channel selection optimization sub-module is used for evaluating the effectiveness, scientificity and rationality of a plurality of channels based on a frequency adjustment scheme by utilizing an analysis hierarchical process, optimizing information release channels, analyzing the plurality of release channels by adopting a genetic algorithm, matching an optimal combination and generating an information release strategy.

The traffic information release method is executed based on a traffic information release system and comprises the following steps:

s1: based on the collected traffic information, adopting a multi-attribute decision making technology to quantitatively score the urgency degree, the influence range and the time sensitivity attribute of the information, comprehensively scoring by a weighting and sequencing method, distributing the priority of each piece of information, and generating an information priority score;

S2: based on the information priority score, analyzing user behavior data by adopting an isolated forest algorithm, identifying an abnormal behavior mode, subdividing the abnormal mode by a density clustering algorithm, identifying an atypical query mode and a demand mutation, and generating a user behavior abnormal index;

s3: based on the abnormal user behavior index, an autoregressive moving average model and a circulating neural network are adopted to analyze historical traffic data trend, integrate cross-domain data, including traffic monitoring, weather and social activities, predict traffic flow and congestion situation, and generate traffic trend prediction analysis;

S4: based on the traffic trend prediction analysis, adopting a graph convolution network to analyze the relevance between a plurality of nodes and paths in the traffic network, predicting potential congestion points and influence on surrounding traffic flows, and generating traffic flow relevance evaluation;

S5: based on the traffic flow relevance evaluation and the user behavior abnormality index, a convolutional neural network and a cyclic neural network are adopted to analyze the user traveling habit, and a personalized traveling scheme is provided by combining the real-time traffic condition to generate personalized traffic information recommendation;

s6: based on the personalized traffic information recommendation, adopting a dynamic adjustment algorithm to optimally adjust release contents, frequencies and channels through real-time traffic conditions and user feedback, and generating optimized release contents and frequencies;

S7: based on the optimized release content and frequency, a dynamic adjustment algorithm is adopted, and through the characteristics of various channels and user preferences, proper information release channels are matched, so that an information release strategy is generated.

A computer device comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing a traffic information distribution system as described above when executing the computer program.

A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of the traffic information distribution method as described above.

Compared with the prior art, the invention has the advantages and positive effects that:

In the invention, the traffic information is comprehensively scored by adopting the multi-attribute decision-making technology, so that the accurate judgment of the information priority is realized, the urgent and important traffic information is issued preferentially, and the real-time performance and accuracy of traffic management are effectively improved. The user behavior is accurately analyzed through the isolated forest algorithm and the clustering algorithm, the abnormal behavior mode is identified, and the response is quick, so that the user behavior analysis mechanism greatly improves the response capability and the prediction accuracy of the system to the emergency. And integrating and analyzing large-range cross-domain data through an autoregressive moving average model and a circulating neural network, so that the prediction of traffic flow and congestion is more comprehensive and accurate. By the application of the graph rolling network, the system is allowed to predict potential congestion points and evaluate the influence of the potential congestion points on surrounding traffic flows, and more scientific and accurate basis is provided for traffic planning and management. Personalized traffic information recommendation is provided through analysis results, user experience and satisfaction are greatly improved, content, frequency and channels of information release are optimized through a dynamic adjustment algorithm, timeliness and relevance of information release are ensured, and quality and efficiency of information service are effectively improved.

Drawings

FIG. 1 is a system flow diagram of the present invention;

FIG. 2 is a system block diagram of the present invention;

FIG. 3 is a flow chart of an information evaluation module according to the present invention;

FIG. 4 is a flowchart of a user behavior analysis module according to the present invention;

FIG. 5 is a flow chart of a cross-domain data analysis module according to the present invention;

FIG. 6 is a flowchart of a correlation analysis module according to the present invention;

FIG. 7 is a flowchart of an information recommendation module according to the present invention;

FIG. 8 is a flowchart of a publication policy optimization module of the present invention;

fig. 9 is a flow chart of method steps of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.

In the description of the present invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate orientations or positional relationships based on the orientation or positional relationships shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and therefore should not be construed as limiting the present invention. Furthermore, in the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.

Example 1

Referring to fig. 1, the present invention provides a technical solution: the traffic information issuing system comprises an information evaluation module, a user behavior analysis module, a cross-domain data analysis module, a correlation analysis module, an information recommendation module and an issuing strategy optimization module;

The information evaluation module adopts a multi-attribute decision making technology to quantitatively score the urgency degree, the influence range and the time sensitivity attribute of the information based on the collected traffic information, adopts a weighting and sorting method to carry out weight distribution and comprehensive scoring on the multi-item attribute, sorts each piece of information, optimizes the sorting result through a technical sorting method, determines the information priority, converts the multi-dimensional information attribute into a single comprehensive score, and generates an information priority evaluation result;

The user behavior analysis module analyzes user behavior data by adopting an isolated forest algorithm based on the information priority evaluation result, identifies abnormal behaviors, subdivides the abnormal behavior patterns by adopting a clustering algorithm, identifies atypical query patterns and demand mutation, refines information, and generates a user behavior abnormal analysis result;

The cross-domain data analysis module processes historical traffic data by adopting an autoregressive moving average model based on a user behavior anomaly analysis result, analyzes trend and seasonal change, predicts traffic flow, integrates traffic monitoring, weather and social activity cross-domain data by adopting a cyclic neural network technology, predicts traffic flow and congestion situation by deep learning, analyzes and extracts cross-domain data correlation, and generates a traffic prediction analysis scheme;

the information recommendation module analyzes the travel habit and preference of the user and learns the travel mode by adopting a convolutional neural network based on the traffic flow relevance analysis result and the user behavior abnormality analysis result, processes real-time traffic condition data by utilizing the convolutional neural network, predicts traffic condition change in a short period, analyzes the user requirements and the current traffic condition and generates a personalized traffic information recommendation result;

The issuing strategy optimization module optimizes the content, frequency and channel of information issuing through a dynamic adjustment algorithm based on personalized traffic information recommendation results, analyzes user feedback and real-time traffic condition change, adjusts the issuing strategy to match current requirements, optimizes timeliness and relevance of information issuing, and generates an information issuing strategy;

The information priority evaluation result comprises an emergency degree index, an influence area size and an expected influence duration, the user behavior abnormality analysis result comprises an abnormality query frequency, an abnormality query time period and atypical route selection, the traffic prediction analysis scheme comprises predicted traffic flow change, a potential congestion area and key external factors affecting traffic flow, the traffic flow relevance analysis result comprises relevance scores of key nodes and paths, predicted congestion nodes and traffic flow transmission paths, the personalized traffic information recommendation result comprises a user preference route, a suggested travel time period and a coping strategy, and the information release strategy comprises information content adjustment parameters, an adjustment rule of release frequency and a selected release channel.

In the information evaluation module, the collected traffic information is firstly formatted and converted into a structured data format which can be processed by an algorithm, a multi-attribute decision making technology is adopted, the emergency degree, the influence range and the time sensitivity attribute of each piece of traffic information are quantitatively scored by adopting a weighting and sequencing method and a technical sequencing method, the attribute is given with weight, the information is scored, the scoring of each attribute is integrated, the ranking is carried out, the weight among various attributes is reasonably distributed, and the technical sequencing method provides an effective mode for determining the priority of each information item on the basis of considering the optimal solution and the worst solution. Accurate assessment of the information priority is ensured, the generated information priority assessment result is convenient for reference of a subsequent module, and the multidimensional information attribute is effectively converted into a single comprehensive score to guide a subsequent decision process.

And in the user behavior analysis module, based on the information priority evaluation result, analyzing the query frequency and the query time behavior data of the user by adopting an isolated forest algorithm, and identifying an abnormal behavior mode which is obviously deviated from a normal state. And refining the abnormal behavior through a density clustering algorithm, and distinguishing atypical query patterns and demand mutation. The isolated forest algorithm effectively isolates abnormal points by constructing a random forest, and the density clustering algorithm clusters according to density relativity among data points to identify various abnormal behavior groups. The system can accurately capture and analyze the abnormal mode in the user behavior, so that a user behavior abnormal analysis result is generated, and accurate data support is provided for personalized recommendation and abnormal behavior monitoring.

In the cross-domain data analysis module, the system integrates the analysis result of the abnormal user behavior and the historical traffic data, and adopts an autoregressive moving average model and a circulating neural network to conduct data analysis. And analyzing historical trend and seasonal change of traffic flow through a moving average model, analyzing time sequence data and cross-domain data through a cyclic neural network, and carrying out deep learning to predict traffic flow and congestion in a short period by integrating traffic monitoring, weather and social activities. By adopting a cross-domain data integration and depth analysis method, valuable information is extracted from a wide data source through a system, a traffic prediction analysis scheme is generated, and a comprehensive decision basis is provided for traffic management and planning.

In the association analysis module, based on the traffic prediction analysis scheme, the system analyzes the association between nodes and paths of the traffic network through the graph rolling network. The graph rolling network learns the dependency relationship among nodes while maintaining the node structure information, effectively predicts potential congestion points and the influence of the potential congestion points on surrounding traffic flows, reveals the distribution characteristics and dynamic change rules of the traffic flows, generates a traffic flow relevance analysis result, provides a more accurate prediction tool for traffic management, and enables congestion prevention and traffic scheduling to be more scientific and efficient.

In the information recommendation module, a convolutional neural network and a cyclic neural network are adopted by the system in combination with a traffic flow relevance analysis result and a user behavior abnormality analysis result, and travel habits and preferences of users are analyzed. The convolutional neural network is used for analyzing historical trip data of a user, identifying common routes and preference time, processing real-time traffic condition data by the convolutional neural network and predicting short-term changes of traffic conditions. The system can generate personalized traffic information recommendation, meets personalized requirements of users, and improves efficiency and comfort of travel decision.

In the issuing strategy optimization module, based on personalized traffic information recommendation results, the system adopts a dynamic adjustment algorithm to comprehensively optimize issuing contents, frequency and channels. The system constantly analyzes user feedback and real-time traffic condition changes, dynamically adjusts information release strategies, and ensures timeliness, relevance and user participation of the information. The dynamic adjustment algorithm enables the release strategy to always keep an optimal state, and the generated information release strategy aims at ensuring effective transmission of traffic information and coverage of the maximum range, so that the overall efficiency and user satisfaction of a public transportation system are improved.

Referring to fig. 2 and 3, the information evaluation module includes an emergency degree analysis sub-module, an influence range analysis sub-module, and a time sensitivity analysis sub-module;

The emergency degree analysis submodule quantitatively scores the emergency degree of the information by adopting a multi-attribute decision making technology based on the collected traffic information, and converts the emergency condition of the information into a numerical score by setting an emergency degree quantization standard to generate an emergency degree score;

The influence range analysis submodule carries out quantitative scoring on the influence range of the information by adopting a weighting and sorting method based on the emergency degree scoring, carries out weight distribution by the emergency degree scoring, and generates an influence range scoring;

The time sensitivity analysis submodule adopts a technical sequencing method to quantitatively score the time sensitivity of the information based on the influence range score, comprehensively analyzes the emergency degree score and the influence range score to determine the priority of the information and generate an information priority assessment result;

In the emergency analysis sub-module, the system first converts the collected traffic information into a unified data format for processing. The module quantitatively scores the emergency degree of each piece of traffic information through a multi-attribute decision making technology. By setting a set of quantification standards including accident type, number of affected lanes and estimated cleaning time, each factor is given different weights according to the degree of influence on traffic flow. Then, the system scores each piece of information according to the quantization standard, and accumulates each weighted score to generate a final emergency degree score. The score reflects the emergency influence degree of each traffic event on the system, a quantitative basis is provided for subsequent decisions, subjective judgment is effectively converted into objective data, and the efficiency and accuracy of information processing are improved.

In the influence range analysis sub-module, the system evaluates the influence range of the traffic event based on the urgency score obtained in the previous step. And (3) quantitatively scoring the influence ranges of different events by adopting a weighted and sequencing method, wherein the quantitative scoring comprises considering the size of a geographic area influenced by the events, the quantity of influenced traffic and the chain reaction of the peripheral area. The quantization process herein involves weighting a number of factors according to their degree of influence, and summing the weighted values to obtain a final influence range score. The space influence of traffic events is quantified, and an administrator can more accurately identify and prioritize events with a wide influence range, and more effectively allocate resources and schedule emergency response.

In the time sensitivity analysis submodule, the system comprehensively considers the emergency degree score and the influence range score, and adopts a technical sorting method to quantitatively score the time sensitivity of the traffic information. A plurality of time-related factors such as the time at which the event occurred, the expected duration, and its impact on the peak traffic period are analyzed. By comprehensive evaluation of the target time factors, the system can determine the time sensitivity score of each event and determine the priority of event processing. By analyzing potential development trend and influence on future traffic flow based on the current state of the event, a more comprehensive and prospective decision basis is provided for traffic management departments, so that the most effective measures can be taken at key moments, and the long-term influence of traffic jams and accidents is reduced.

Referring to fig. 2 and 4, the user behavior analysis module includes a query frequency analysis sub-module, a query time analysis sub-module, and a route selection analysis sub-module;

The query frequency analysis submodule analyzes the query frequency of the user by adopting an isolated forest algorithm based on the information priority evaluation result, screens the query behavior of the abnormal frequency and generates a query frequency abnormal index;

The query time analysis submodule analyzes the time distribution of the query behaviors of the user by adopting a clustering algorithm based on the abnormal index of the query frequency, identifies abnormal query behaviors in atypical time periods, identifies the effect of time factors on the abnormal behaviors of the user and generates an abnormal index of the query time;

The route selection analysis sub-module analyzes the route selection behavior of the user by adopting an isolated forest algorithm based on the abnormal index of the query time, identifies the atypical query mode of the user, evaluates the abnormal degree of the behavior of the user by the query frequency of the user and the abnormal index of the query time, and generates an abnormal analysis result of the behavior of the user;

In the query frequency analysis sub-module, the system analyzes the query frequency of the user by using an isolated forest algorithm according to the information priority evaluation result. The data is first formatted into a structure that is easy for the algorithm to process, including user ID, query time stamp, and query content. The isolated forest algorithm isolates data points by constructing a plurality of random trees, distinguishing frequently queried user behavior from normal query behavior. In the execution process, the algorithm evaluates the difficulty level of each data point to be isolated and judges whether the data points are abnormal or not. The generated abnormal index of the query frequency clearly indicates which users exhibit atypical high-frequency query behaviors, which plays an important role in identifying potential information security problems, system abuse or demand prediction and the like.

In the query time analysis sub-module, the system analyzes the time distribution of the user query behavior by using a clustering algorithm. Clustering query behaviors according to time distribution through query frequency abnormal indexes generated by a previous submodule, and identifying whether query behaviors which occur in a concentrated mode in a target time period are abnormal or not. The clustering algorithm is used to group query times, identifying which query actions are abnormally frequent during an irregular time period. This analysis considers periodic and seasonal patterns of user behavior, revealing query behavior that does not conform to the conventional pattern. The generated query time anomaly index provides time attributes for the system to be informed of user behaviors, is beneficial to improving information service and user experience, and has a key effect on monitoring and preventing improper use of the system.

In the route selection analysis sub-module, the system adopts an isolated forest algorithm to analyze the route selection behavior of the user. Based on the query time anomaly metrics, it is identified which users exhibit atypical patterns in selecting routes, such as frequently selecting non-optimal routes or repeatedly querying the same unusual route at different times. Focusing on the query frequency and time of the user and the specific content of the query, including the route selected by the user. By analyzing the anomaly pattern of user route selection, the system can evaluate the degree of anomalies in user behavior, identifying potential data anomalies or changes in user demand. The generated user behavior anomaly analysis result not only provides deep user behavior insight for traffic management, but also provides a basis for subsequent personalized recommendation and service optimization.

Referring to fig. 2 and 5, the cross-domain data analysis module includes a traffic flow prediction sub-module, a congestion point identification result sub-module, and an external factor influence analysis sub-module;

the congestion point identification result submodule is used for integrating real-time traffic monitoring data by adopting a cyclic neural network technology based on traffic flow trend prediction, analyzing and predicting congestion points, processing time sequence data, identifying areas and time of flow sudden increase and generating a congestion point identification result;

The external factor influence analysis submodule integrates cross-domain data based on a congestion point identification result by adopting a cyclic neural network technology, analyzes weather and social activities affecting traffic flow and congestion conditions, predicts the influence of the cross-domain data on the traffic flow by a deep learning model, and generates a traffic prediction analysis scheme;

In the traffic flow prediction sub-module, the system processes and analyzes historical traffic data by adopting an autoregressive moving average model based on the analysis result of the user behavior abnormality. To collecting and collating historical traffic flow data, including traffic flow, time, weather conditions, etc., and formatting such data into a time series data format that can be processed by a model. The autoregressive moving average model is a statistical model for understanding the variation trend and seasonal factor of time series data, and predicting traffic flow in a future period of time by analyzing patterns and periodicity in historical data. Future traffic flow is predicted by identifying and utilizing trends and seasonal variations in the historical data to build a predictive model. The generated traffic flow trend prediction provides reliable data support for traffic management departments, and assists the traffic management departments in understanding traffic conditions which may occur in the future, so that planning and scheduling are performed in advance, traffic flows are optimized, and congestion is reduced.

In the congestion point identification result sub-module, the system integrates and analyzes real-time traffic monitoring data by adopting a cyclic neural network technology based on traffic flow trend prediction results. The cyclic neural network is a deep learning model specially processing sequence data, and can capture the dynamic characteristics and long-term dependency relationship of time sequence data. The method comprises the steps of collecting real-time traffic data, including vehicle speed, vehicle flow density and the like, comparing target data with historical trends, analyzing the target real-time data through a cyclic neural network model, and identifying modes and trends which possibly cause traffic jam. The location and time of impending congestion is identified by analyzing outlier data points that do not correspond to predicted trends, including areas of sudden increases in traffic. The generated congestion point identification result can help traffic management departments respond to the congestion situation which occurs immediately in real time, and preventive measures are taken, including adjustment of traffic signals or recommendation of issuing routes, so that the congestion degree is relieved.

In the external factor influence analysis submodule, a cyclic neural network technology is adopted to analyze cross-domain data, wherein factors such as meteorological conditions, social activities, holiday arrangement and the like can influence traffic flow and congestion conditions. And identifying the relevance between the external factors and the traffic flow through a deep learning model, and predicting the potential influence on the future traffic condition. The target multi-source data are fused together through a cyclic neural network, and the internal relation and the mode of the data are learned. The individual influence of each factor and how the interactions between the target factors affect the traffic flow are analyzed. The generated traffic prediction analysis scheme comprises comprehensive prediction of future traffic conditions, so that traffic management and planning are more comprehensive and prospective, and complex and changeable traffic conditions can be effectively dealt with.

Referring to fig. 2 and fig. 6, the association analysis module includes a node association evaluation sub-module, a congestion prediction sub-module, and a traffic propagation analysis sub-module;

The congestion prediction sub-module adopts a long-period memory network to evaluate potential congestion points based on node relevance scores, captures long-period dependency relations in time sequence data, analyzes traffic flow data and node relevance, predicts congestion conditions occurring in a target time period, and generates congestion point prediction information;

The traffic flow propagation analysis submodule analyzes the propagation modes of traffic flow, including the relevance among nodes and traffic flow change, based on the congestion point prediction information, applies an infectious disease propagation model, analyzes the distribution characteristics and dynamic change rules of traffic flow in the whole network and generates a traffic flow relevance analysis result;

In the node relevance evaluation sub-module, based on a traffic prediction analysis scheme, nodes and paths in a traffic network are deeply analyzed by using a graph rolling network technology. By converting the traffic network into a graph structure, wherein nodes represent traffic intersections or sections, edges represent road connections, and adopting graph convolution network technology, the relevance of each node and surrounding nodes is analyzed, including the dependence of traffic flow and the influence of paths. The graph rolling network can capture complex interactions among nodes, and the influence degree of different nodes on the whole traffic network is evaluated by analyzing the interaction of target nodes, so that node relevance scores are given. The generated node relevance scores are helpful for traffic managers to identify key nodes and fragile paths, and scientific basis is provided for optimizing traffic flow and relieving congestion.

In the congestion prediction submodule, the system analyzes the interaction among nodes and potential congestion points by adopting a graph rolling network technology based on node relevance scores. By combining node scoring with real-time traffic data, it is predicted which nodes are likely to become future congestion hotspots. Through the graph rolling network, static attributes of the nodes including geographic positions, road types and dynamic factors including real-time traffic flow and event influence are analyzed. By deeply analyzing interactions between nodes, the system is able to identify nodes and paths that are potentially congested due to traffic flow fluctuations. The generated congestion point prediction information result can guide a traffic management department to conduct timely traffic scheduling and route planning, and congestion is effectively prevented or relieved.

In the traffic propagation analysis sub-module, the system continues to utilize graph convolution network technology to analyze how traffic flows propagate throughout the network based on congestion point prediction information results. By simulating a variety of traffic conditions, including accidents, road closures, or large activities, it is analyzed how the target event affects the traffic flow of surrounding nodes and paths. The graph rolling network can analyze the global structure and local details of the traffic network in the link and predict the propagation mode and the influence range of traffic flow. By simulating traffic propagation under different conditions, the system can predict the distribution characteristics and dynamic changes of traffic flow and generate traffic flow relevance analysis results. Providing deep insight for traffic planning and management, helping managers understand and predict the change trend of traffic flow, and providing basis for implementing effective traffic control and emergency response measures.

Referring to fig. 2 and 7, the information recommendation module includes a travel habit analysis sub-module, a real-time status analysis sub-module, and a recommendation policy generation sub-module;

The travel habit analysis submodule analyzes user history travel data based on the traffic flow relevance analysis result and the user behavior abnormality analysis result by adopting a convolutional neural network technology, identifies user travel habits and preferences, and generates a user travel habit analysis result through a user history travel mode comprising a common route and travel time;

The recommendation strategy generation submodule is used for analyzing travel time and route by using a convolutional neural network and a cyclic neural network technology according to the real-time traffic condition analysis result and through user travel habits and current traffic conditions, and generating a personalized traffic information recommendation result;

In the trip habit analysis submodule, the system adopts a convolutional neural network technology to deeply analyze the historical trip data of the user based on the traffic flow relevance analysis result and the user behavior abnormality analysis result. And converting travel records of the user, including information such as route selection, travel time, frequency and the like, into a format suitable for processing by a neural network, including time series or imaging data. And carrying out feature extraction and pattern recognition on the data by using a convolutional neural network, and recognizing the travel habits and preferences of the user, such as preferred travel routes, travel time periods and the like. The convolutional neural network learns deep features of trip data through a multi-layer filter and captures a behavior pattern of a user. The generated user travel habit analysis result provides a basis for subsequent real-time traffic condition analysis and personalized recommendation, so that a recommendation system can provide more accurate service according to specific requirements and historical behaviors of a user.

In the real-time condition analysis sub-module, the system analyzes the real-time traffic condition data by using a cyclic neural network technology based on the analysis result of the travel habit of the user. Including collecting current traffic information such as road congestion conditions, accident situations, weather conditions, etc., and formatting the information into serial data. The recurrent neural network is particularly suitable for processing such sequence data, and can identify the dynamic change of the current traffic condition by referring to the time sequence characteristic of the data. By combining the travel habits of the users with the real-time traffic information, the system can evaluate the influence of the traffic conditions on the travel of the users in real time, and generate real-time traffic condition analysis results. The method is crucial for updating the travel advice of the user in real time, helps the user avoid congestion, and optimizes the travel route.

In the recommendation strategy generation sub-module, the system performs deep analysis of travel time and routes by combining travel habits of users and current traffic conditions by utilizing a convolutional neural network and a cyclic neural network technology again based on real-time traffic condition analysis results. The convolutional neural network extracts spatial features of the line data, including geographical information of routes, traffic density and the like, and the convolutional neural network analyzes temporal features of the data, including time periods of travel and time sequences of traffic states. By the method for using the convolutional neural network and the cyclic neural network in a mixed mode, various factors can be comprehensively considered by the system, and personalized travel suggestions can be generated for the user. The generated personalized traffic information recommendation result considers the personal preference of the user and the real-time traffic condition, so that the traveling efficiency and the traveling comfort are improved.

Referring to fig. 2 and 8, the distribution policy optimization module includes a content adjustment policy sub-module, a distribution frequency management sub-module, and a channel selection optimization sub-module;

The content adjustment strategy sub-module extracts user characteristics by adopting a text mining algorithm based on personalized traffic information recommendation results, classifies and prioritizes information by utilizing a support vector machine algorithm through historical preference and feedback of a user, analyzes information content, performs personalized adjustment and generates a content adjustment scheme;

the release frequency management submodule is based on a content adjustment scheme, adopts time sequence analysis, matches information release time points, adjusts information release frequency, and generates a frequency adjustment scheme by adjusting release frequency through a prediction result and real-time traffic conditions by using a Poisson distribution model;

the channel selection optimization sub-module is used for evaluating the effectiveness, scientificity and rationality of a plurality of channels based on a frequency adjustment scheme by utilizing an analysis layering process, optimizing information release channels, analyzing the plurality of release channels by adopting a genetic algorithm, and matching an optimal combination to generate an information release strategy;

In the content adjustment strategy sub-module, the system adopts a dynamic adjustment algorithm to analyze the current traffic condition and the user demand based on the personalized traffic information recommendation result. By collecting and integrating current traffic condition data, including traffic flow, accident information, road engineering information, etc., and user feedback and preference information. By formatting the data into a standard format for algorithmic processing. And carrying out real-time adjustment on the released traffic information content according to the real-time data and the user preference through a dynamic adjustment algorithm, wherein the real-time adjustment comprises the steps of updating road condition information, adjusting recommended routes and the like. Including real-time analysis and decision-making of data, ensures timeliness and relevance of information content. The generated content adjustment scheme reflects the latest traffic condition and considers the personalized requirements of the users, thereby improving the user satisfaction and the practicability of the information service.

In the release frequency management sub-module, the system adopts a dynamic adjustment algorithm to optimize the information release frequency based on a content adjustment scheme. By evaluating the speed of change of the current traffic condition and the frequency of the user's demand for information. The dynamic adjustment algorithm analyzes the target factors and determines the optimal frequency of issuing information, including increasing the issue frequency during peak hours and decreasing the issue frequency during night or traffic stationary hours. The information release frequency can meet the requirements of users and can not cause information overload. By optimizing the release frequency, the system can prevent users from feeling information flooding while ensuring timely information updating, thereby improving user experience and information utility.

In the channel selection optimization sub-module, the system analyzes and selects the optimal information distribution channel by adopting a dynamic adjustment algorithm based on a frequency adjustment scheme. By examining the coverage efficiency and user preference of different channels, including social media, mobile application notification, electronic screen and the like, as well as the real-time effect and user arrival rate of each channel, the dynamic adjustment algorithm selects an appropriate channel for information release according to the analysis result, so that information can be efficiently and accurately transmitted to a target user group. By optimizing channel selection, the efficiency and effectiveness of information transmission are improved, and key information can be ensured to reach the hand of a user in need of the key information, so that the response speed and efficiency of the whole traffic information service system are improved.

Referring to fig. 9, the traffic information issuing method includes the steps of:

S2: based on the information priority score, analyzing user behavior data by adopting an isolated forest algorithm, identifying an abnormal behavior mode, subdividing the abnormal mode by adopting a density clustering algorithm, identifying an atypical query mode and a demand mutation, and generating a user behavior abnormal index;

S3: based on abnormal user behavior indexes, an autoregressive moving average model and a circulating neural network are adopted to analyze historical traffic data trend, integrate cross-domain data, including traffic monitoring, weather and social activities, predict traffic flow and congestion situation, and generate traffic trend prediction analysis;

S4: based on traffic trend prediction analysis, adopting a graph rolling network to analyze the relevance between a plurality of nodes and paths in a traffic network, predicting potential congestion points and influence on surrounding traffic flows, and generating traffic flow relevance evaluation;

S6: based on personalized traffic information recommendation, adopting a dynamic adjustment algorithm, and optimizing and adjusting release contents, frequencies and channels through real-time traffic conditions and user feedback to generate optimized release contents and frequencies;

S7: based on the optimized release content and frequency, adopting a dynamic adjustment algorithm, and matching proper information release channels through the characteristics of various channels and user preferences to generate an information release strategy;

The collected traffic information is quantitatively scored by adopting a multi-attribute decision making technology, so that the processing of the traffic information is ensured to be more accurate and objective. By adopting a weighting and sequencing method, the quantitative scoring and the comprehensive scoring of the attributes such as the emergency degree, the influence range, the time sensitivity and the like ensure that the priority of each piece of information is accurately distributed. The information processing mode improves the efficiency of traffic information management, enables emergency situations to be responded and processed more quickly, reduces risks of traffic jam and accidents, and improves the road use efficiency. The user behavior data is analyzed through an isolated forest algorithm, the abnormal behavior mode is identified, the abnormal mode is subdivided through a density clustering algorithm, atypical inquiry modes and demand mutation are identified, and a deeper user behavior analysis result is provided for traffic management. The traffic management department is facilitated to know the actual demands of users, potential traffic problems are predicted and prevented to a certain extent, and support is provided for formulating more humanized and efficient traffic scheduling strategies. By combining an autoregressive moving average model and a cyclic neural network to analyze historical traffic data trend and integrate cross-domain data, such as traffic monitoring, weather, social activities and the like, traffic flow and congestion conditions can be comprehensively predicted. The comprehensive prediction enables the traffic trend prediction to be more accurate and comprehensive by being based on historical data and various external factors influencing traffic flow. The accurate prediction result can help traffic management departments to prepare in advance, effectively relieve or avoid congestion, and improve the overall operation efficiency of the traffic system. And analyzing the relevance between a plurality of nodes and paths in the traffic network by adopting the graph convolution network, accurately predicting potential congestion points, and evaluating the influence of the congestion points on surrounding traffic flows. The traffic flow can be optimized on the macroscopic level and the microscopic level, a more detailed and deep analysis result is provided for traffic management departments, and more effective traffic scheduling and control measures are facilitated to be formulated, so that traffic delay is reduced, and road traffic capacity is improved. By combining the convolutional neural network and the cyclic neural network technology, a personalized trip scheme is provided, and the release content, the frequency and the channels are optimally adjusted through a dynamic adjustment algorithm, so that trip requirements of different users can be met, and timeliness and effectiveness of traffic information service can be ensured. The personalized trip scheme enables each user to obtain traffic information suitable for the user, and the optimized release content and frequency ensure the accuracy and high efficiency of information transmission. By matching with proper information release channels, the information can be ensured to be covered to the most extensive users, so that the overall service level and user satisfaction of the public transportation system are improved.

A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the traffic information release system when executing the computer program.

A computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the traffic information distribution method as described above.

The present invention is not limited to the above embodiments, and any equivalent embodiments which can be changed or modified by the technical disclosure described above can be applied to other fields, but any simple modification, equivalent changes and modification made to the above embodiments according to the technical matter of the present invention will still fall within the scope of the technical disclosure.

Claims

1. The traffic information issuing system is characterized by comprising an information evaluation module, a user behavior analysis module, a cross-domain data analysis module, a correlation analysis module, an information recommendation module and an issuing strategy optimization module;

2. The traffic information distribution system according to claim 1, wherein the information evaluation module includes an emergency degree analysis sub-module, an influence range analysis sub-module, a time sensitivity analysis sub-module;

3. The traffic information issuing system according to claim 1, characterized in that the user behavior analysis module comprises a query frequency analysis sub-module, a query time analysis sub-module, a route selection analysis sub-module;

4. The traffic information issuing system according to claim 1, characterized in that the cross-domain data analysis module comprises a traffic flow prediction sub-module, a congestion point identification result sub-module, and an external factor influence analysis sub-module;

5. The traffic information issuing system according to claim 1, characterized in that the association analysis module comprises a node association evaluation sub-module, a congestion prediction sub-module and a traffic propagation analysis sub-module;

6. The traffic information issuing system according to claim 1, characterized in that the information recommending module comprises a travel habit analyzing sub-module, a real-time condition analyzing sub-module and a recommending strategy generating sub-module;

7. The traffic information issuing system according to claim 1, characterized in that the issuing policy optimization module comprises a content adjustment policy sub-module, an issuing frequency management sub-module, and a channel selection optimization sub-module;

8. A traffic information distribution method, characterized in that the traffic information distribution method is executed based on the traffic information distribution system according to any one of claims 1 to 7, comprising the steps of:

based on the collected traffic information, adopting a multi-attribute decision making technology to quantitatively score the urgency degree, the influence range and the time sensitivity attribute of the information, comprehensively scoring by a weighting and sequencing method, distributing the priority of each piece of information, and generating an information priority score;

based on the information priority score, analyzing user behavior data by adopting an isolated forest algorithm, identifying an abnormal behavior mode, subdividing the abnormal mode by a density clustering algorithm, identifying an atypical query mode and a demand mutation, and generating a user behavior abnormal index;

Based on the abnormal user behavior index, an autoregressive moving average model and a circulating neural network are adopted to analyze historical traffic data trend, integrate cross-domain data, including traffic monitoring, weather and social activities, predict traffic flow and congestion situation, and generate traffic trend prediction analysis;

Based on the traffic trend prediction analysis, adopting a graph convolution network to analyze the relevance between a plurality of nodes and paths in the traffic network, predicting potential congestion points and influence on surrounding traffic flows, and generating traffic flow relevance evaluation;

Based on the traffic flow relevance evaluation and the user behavior abnormality index, a convolutional neural network and a cyclic neural network are adopted to analyze the user traveling habit, and a personalized traveling scheme is provided by combining the real-time traffic condition to generate personalized traffic information recommendation;

based on the personalized traffic information recommendation, adopting a dynamic adjustment algorithm to optimally adjust release contents, frequencies and channels through real-time traffic conditions and user feedback, and generating optimized release contents and frequencies;

based on the optimized release content and frequency, a dynamic adjustment algorithm is adopted, and through the characteristics of various channels and user preferences, proper information release channels are matched, so that an information release strategy is generated.

9. A computer device comprising a memory and a processor, characterized in that the memory has stored therein a computer program, which processor, when executing the computer program, implements the traffic information distribution system according to any of claims 1 to 7.

10. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the traffic information distribution method according to claim 8.