| With the rapid development of China’s economy,the dangerous goods transportation industry has expanded rapidly,causing frequent accidents in the transportation of dangerous goods.And the accident harm and negative social impact are huge.Based on the analysis of the historical data of dangerous goods transportation accidents of the Ministry of Emergency Management of the People’s Republic of China,this paper analyzes the influencing factors of dangerous goods transportation with the current situation of road transportation supervision in Shaanxi Province.This paper studies the real-time risk warning methods of dangerous goods transportation,and constructs the real-time risk evaluation index system of dangerous goods transportation.Based on hybrid deep learning GRUDNN(Gate Recurrent Unit Dynamic Neural Network,GRUDNN)algorithm,the real-time risk warning model for dangerous goods transportation is constructed.Designed and implemented a real-time warning system for dangerous goods transportation risks based on hybrid deep learning GRUDNN.The main work of this paper is as follows:(1)Starting from the four factors of people,vehicles,roads and surroundings,and based on statistics and analysis of dangerous goods transportation accidents.Combined with the results of questionnaires and expert opinions,determine the real-time risk indicators and parameter definitions of dangerous goods transportation.And optimize various indicators,based on which comprehensive assessment of the real-time risk of dangerous goods transportation.The dynamic optimization real-time risk evaluation index system of dangerous goods transportation is constructed to prepare for the real-time risk early warning algorithm of dangerous goods transportation.(2)In view of the lack of accuracy of the existing risk assessment methods for the dynamic assessment of dangerous goods transportation risks,and the flexibility to deal with unexpected events.A real-time risk warning algorithm for dangerous goods transportation based on hybrid deep learning GRUDNN is proposed.Finally,in order to increase the model’s anti-interference ability,the predictive similarity method is used to improve the robustness of the algorithm.(3)The Vue.js framework is used to construct the system interface and My SQL database to manage data,and visualizes the real-time risk of dangerous goods transportation.Four colors of red,blue,green,and yellow are used to indicate the current vehicle risk level.The real-time risk early warning model of dangerous goods transportation based on hybrid deep learning grudnn is used as the main back-end processing algorithm.A real-time risk early warning system of dangerous goods transportation based on hybrid deep learning is formed,which integrates real-time monitoring and risk prediction. |
