Research On RNN-based Traffic Accident Prediction And Rescue Resource Allocating Algorithm

Allocating algorithm In Intelligent Transportation System(ITS),Emergency Rescue System(EMS)is a special system,which is based on other subsystems of Intelligent Transportation System(ITS),linking traffic monitoring center and professional rescue agencies into an organic whole.In our daily life,the occurrence of traffic accidents is often unpredictable,which requires the rapid response of a sound emergency rescue system to reduce the loss of accidents.Generally speaking,when an accident occurs,a lot of disaster relief resources will be needed to deal with the situation at the same time,and the disaster relief resources may not be inexhaustible.How to achieve the most efficient resource allocating in the shortest time has become one of the concerns of researchers.In view of the above problems,we propose the solutions which are shown as follows:(1)RNN-based forecast algorithm for traffic emergency situationThis paper proposes a traffic accident prediction algorithm based on RNN model.The algorithm processes the real-time data uploaded by the vehicle in the Internet of Vehicles and the information collected by the network infrastructure,and trains a prediction model based on the historical data information.The model analyzes the road information in real time according to the existing traffic or weather conditions,and uses the RNN model to predict the accident probability,which provides a basis for subsequent resource allocating.It can be seen from the training results of the model that the accuracy and misjudgment of the prediction meet the requirements of ITS operation,which can be used as an important reference for resource allocating in ITS.(2)Prediction based resource allocating algorithm for ITSSecondly,the resources are pre-allocated according to the severity of the accident predicted by the RNN prediction algorithm.For emergency resources,the system management center can assign it to the predicted accident location in advance to control the traffic command situation.Ideally,it is hopeless for emergency.The presence of resources can alleviate traffic conditions and avoid accidents.For other resources,if the system makes predictions and judgments,if the accidents are not successfully avoided,resources can be allocated at the first time of the accident,saving the emergency situation.The time it takes for resources to be allocated cooperatively.The disaster relief efficiency of different types of accidents is also different according to the pre-allocation of resources.The experiments in this paper prove that the time spent by the algorithm based on the predicted resource pre-allocation is shorter than the time spent by the rescue system which is operated after the accident.At the same time,it can be found from the comparison of experimental results that the proposed resource allocating algorithm based on prediction has a substantial improvement on the efficiency of accident processing compared with the traditional algorithm.