Urban traffic accident black spot treatment method and system based on artificial intelligence

Abstract

The invention discloses an urban traffic accident black spot treatment method and system based on artificial intelligence, and the method specifically comprises the steps: collecting urban traffic accident data through a remote alarm system, storing the data in a platform database, and carrying out the cleaning and standardization processing of the data; using an AI model to identify black spot features and generating a governance strategy; implementing a treatment scheme and feeding back effect data to optimize the AI model; the method and system have the advantages that the black spot recognition efficiency can be improved, and the manual intervention cost is reduced; a governance strategy is dynamically recommended through an AI model, and the scheme suitability and the implementation effect are improved; continuous optimization of a treatment scheme is supported, and intelligent upgrading of traffic management is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent traffic management, and particularly to a method and system for governing black spots of urban traffic accidents based on artificial intelligence. Background Art

[0002] Traditional identification of traffic accident black spots mainly relies on manual experience to analyze historical accident data, with low efficiency and being easily affected by subjective factors. In the prior art, although some traffic processing systems can achieve data visualization, they lack in-depth mining of accident characteristics, and the recommended governance solutions rely on a fixed rule base, making it difficult to dynamically adapt to the specific characteristics of different black spots. In addition, the design and implementation process of black spot governance solutions lack intelligent support and cannot optimize strategies by combining national standards and actual governance effects.

[0003] Chinese invention patent CN202110255857.8 discloses a method for identifying accident black spots based on traffic accident big data, including: obtaining road network data and traffic accident data in a preset time period; performing position matching on the traffic accident data and the road network data; identifying accident black spots, and this step includes: extracting accident high-incidence points; based on the position information of the accident high-incidence points, using the K-means clustering method to cluster the accident high-incidence points; for each obtained cluster, calculating the center point of each cluster respectively; taking each center point as the center, all accident points within a preset radius range form each black spot area. This invention extracts objective regular information such as the spatio-temporal distribution characteristics of traffic accidents based on the already occurred traffic accident data, and can accurately identify traffic accident black spots on the road network. However, this invention does not involve artificial intelligence model training and automatic generation of governance strategies. Therefore, the prior art has the following defects: 1. Insufficient data analysis ability: Unable to efficiently identify the dynamic characteristics of accident black spots; 2. Lack of intelligence in governance strategies: The recommended solutions rely on a static rule base and cannot adapt to complex scenarios; 3. Difficulty in iterative optimization of governance effects: Lack of an artificial intelligence-based governance effect feedback and model training mechanism. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for governing black spots of urban traffic accidents based on artificial intelligence, so as to solve the problems of low efficiency in analyzing accident black spots and insufficient intelligence in governance strategies in the prior art.

[0005] The technical solution adopted by the present invention to solve the above technical problem is: A method for governing black spots of urban traffic accidents based on artificial intelligence, including the following specific steps: (1). Collect urban traffic accident data through a remote alarm system, store it in the platform database, and perform cleaning and standardization processing on the data; (2) Use the AI model to identify the characteristics of black spots and generate governance strategies; (3) Implement the governance plan and feedback the effect data to optimize the AI model.

[0006] Further, before using the AI model to identify black spots, the AI model is trained first. Specifically: Use the historical governance plans and corresponding black spot feature data of urban traffic accident black spots to train the convolutional neural network CNN and the decision tree model, establish the feature-strategy mapping relationship, and set the training parameters. Optimize the model accuracy through cross-validation.

[0007] Further, national standard documents are integrated during the training of the AI model.

[0008] Further, in step (3), the optimization process of the AI model is as follows: Update the training set of the AI model according to the governance effect data and retrain the AI model to improve the accuracy of strategy recommendation.

[0009] A system for implementing the above-mentioned method for governing urban traffic accident black spots based on artificial intelligence includes: A data management module for storing and visualizing traffic accident data and governance plan documents; A data analysis module that uses the AI model to identify the characteristics of accident black spots and generate an analysis report; A black spot governance module that recommends governance strategies based on the AI model and supports the export of plans and feedback on effects.

[0010] Compared with the prior art, the advantages of the present invention are that the black spot recognition efficiency can be improved and the manual intervention cost can be reduced through this governance method and system; and the governance strategies are dynamically recommended through the AI model to improve the adaptability and implementation effect of the plan; support the continuous optimization of the governance plan and promote the intelligent upgrade of traffic management. Specific implementation mode

[0011] The present invention will be further described in detail below with reference to the embodiments.

[0012] Embodiment 1: A method for governing urban traffic accident black spots based on artificial intelligence includes the following specific steps: (1) Collect urban traffic accident data (such as time, location, type, etc.) through a remote alarm system, store it in the platform database, and clean and standardize the data to ensure the data quality input into the AI model; (2) Conduct AI model training, and integrate national standard documents such as GB 5768.2 and GA / T 1567 during training. Specifically: Adopt the historical treatment solutions for urban traffic accident black spots and the corresponding black spot feature data to train a convolutional neural network (CNN) and a decision tree model, establish a feature-strategy mapping relationship, and set training parameters, such as: learning rate of 0.001 and confidence interval of 95%. Then optimize the model accuracy through cross-validation; (3) Use the trained AI model to identify black spot features and generate treatment strategy solutions, such as: adding isolation fences, adjusting signal light timing, etc.; (4) Implement the treatment solution and feedback the effect data. At the same time, update the training set of the AI model according to the treatment effect data and retrain the AI model to optimize the AI model and improve the accuracy of strategy recommendation.

[0013] Embodiment 2: An urban traffic accident black spot treatment system based on artificial intelligence, including: A data management module, mainly including the management of two types of data, specifically: (1) Accident basic data management: Support the storage, editing, export, and visual display of traffic accident data (such as map point marking, heat map generation); (2) Black spot treatment solution management: Archive the electronic files of treatment solutions, extract key strategies in combination with the AI engine, and build a treatment solution library; A data analysis module, which identifies accident black spot features through an AI model and generates an analysis report, specifically: Mine the accident data based on the AI model and output reports on black spot distribution, types, and features; and establish an association model between black spot features and treatment strategies through pattern learning in combination with national standards, such as GB 5768.2, GA / T 1567, etc.; A black spot treatment module, which automatically generates a treatment solution adapted to the black spot features according to the analysis results of the AI model, and supports the export of the solution and the feedback of treatment effect data for iterative optimization of the AI model.

[0014] The protection scope of the present invention includes but is not limited to the above embodiments. The protection scope is subject to the claims. Any replacement, deformation, and improvement that are easily conceivable by those skilled in the art for this technology fall within the protection scope of the present invention.

Claims

1. A method for managing urban traffic accident black spots based on artificial intelligence, characterized in that The specific steps include: (1) Collect urban traffic accident data through the remote alarm system, store it in the platform database, and clean and standardize the data; (2) Use AI models to identify black spot features and generate governance strategies; (3) Implement governance plans and provide feedback on performance data to optimize AI models.

2. The method for managing urban traffic accident black spots based on artificial intelligence as claimed in claim 1, characterized in that: Before using the AI ​​model to identify black spots, the AI ​​model training is performed first, specifically: The historical governance plan for urban traffic accident black spots and the corresponding black spot feature data are used to train the convolutional neural network (CNN) and decision tree model, establish the feature-strategy mapping relationship, set the training parameters, and optimize the model accuracy through cross-validation.

3. The method for managing urban traffic accident black spots based on artificial intelligence as claimed in claim 2, characterized in that: Integrate national standard documents when training AI models.

4. The method for managing urban traffic accident black spots based on artificial intelligence as claimed in claim 2, characterized in that: In the step (3), the optimization process of the AI ​​model is: updating the training set of the AI ​​model according to the governance effect data, and retraining the AI ​​model to improve the accuracy of strategy recommendations.

5. A system for implementing the method for managing urban traffic accident black spots based on artificial intelligence as claimed in claim 1, characterized in that include: Data management module, used to store and visualize traffic accident data and governance plan files; Data analysis module, which uses AI models to identify accident black spot features and generate analysis reports; The black spot governance module recommends governance strategies based on AI models and supports the export of solutions and effect feedback.