The invention discloses a highway tunnel traffic accident prediction method based on a convolutional neural network, and the method comprises the steps: collecting historical accident data, constructing a traffic accident data set, carrying out the data cleaning, and carrying out the data preprocessing, including formatting, abnormal value elimination, vacancy value filling, correlation analysis, normalization processing, oversampling processing, and gray-scale conversion processing. Performing feature extraction to obtain key variables for predicting accident occurrence; constructing a highway tunnel traffic accident prediction model in combination with a convolutional neural network, and performing parallel training; and performing traffic accident prediction by using the trained highway tunnel traffic accident prediction model. According to the method, key features are automatically extracted from a large amount of data by using the CNN, the dependence on professional knowledge and the demand of artificial feature engineering are reduced, the operation cost is reduced, rapid identification and response can be realized, the arrival time of emergency services is shortened, accident consequences are reduced, and the rescue efficiency is improved.
本发明公开了一种基于卷积神经网络的高速公路隧道交通事故预测方法,包括收集历史事故数据,构建交通事故数据集并进行数据清洗、数据预处理,包括格式化、异常值剔除、空缺值填补、相关性分析、归一化处理、过采样处理以及转灰度处理;进行特征提取,得到用于预测事故发生的关键变量;结合卷积神经网络,构建高速公路隧道交通事故预测模型并行训练;利用训练好的高速公路隧道交通事故预测模型进行交通事故预测。本发明利用CNN自动从大量数据中提取关键特征,减少了对专业知识的依赖和人工特征工程的需求,降低了运行成本,能够迅速识别并响应,缩短紧急服务的到达时间,减轻事故后果,提高救援效率。
Highway tunnel traffic accident prediction method based on convolutional neural network
一种基于卷积神经网络的高速公路隧道交通事故预测方法
2024-10-08
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen |
Highway Traffic Accident Detection Method with Fuzzy Neural Network
British Library Conference Proceedings | 2007
|Traffic accident prediction method based on graph convolutional network
European Patent Office | 2023
|Method for analyzing traffic accident risk of highway tunnel group
European Patent Office | 2021
|