The invention relates to a traffic management system based on big data. The traffic management system comprises a big data training subsystem and a dynamic execution subsystem. The richness of the model is increased by introducing a visual angle frame and a variable of a relative visual angle, so that the dependence of a traffic accident prediction strategy on basic data is reduced, meanwhile, correlation prediction of other external parameters related to accident dangerous cases is obtained through big data training, collision prediction simulation is realized, and the prediction efficiency is improved. The traffic dangerous case is predicted through the result of the collision prediction simulation, so that the dangerous case can be predicted and identified even if the camera cannot acquire all images required by identification due to the visual angle problem.
本发明涉及一种基于大数据的交通管理系统,包括大数据训练子系统和动态执行子系统;通过引入视角框以及相对视角的变量增加模型的丰富程度,使交通事故预测的策略对基础数据的依赖减小,而同时通过大数据训练得到外部关于事故险情的其他参数相关性预测,实现碰撞预测模拟,通过碰撞预测模拟的结果预测交通险情,这样即使摄像头由于视角问题不能获取所有识别需要的图像,也可以进行险情的预测和识别。
Traffic management system based on big data
一种基于大数据的交通管理系统
2024-02-02
Patent
Electronic Resource
Chinese
Intelligent traffic integrated management system based on traffic big data
European Patent Office | 2023
|