The System Design And Construction Of Driving Safety Detection And Loss Rate Prediction

With the rapid development of road traffic and the rapid growth of car ownership,the problem of road traffic safety is becoming more and more serious,the road traffic safety problem also gets people’s attention day by day.Therefore,it is of further significance to detection the safety of driving behavior and predict loss rate of driving behavior.Driving behavior can be divided into safe driving behavior and dangerous driving behavior,among which dangerous driving behavior is harmful,so it is necessary to conduct effective detection of dangerous driving behavior.The loss rate reflects the cost and operation status of the insurance company,as well as the road traffic safety status,the road traffic safety status can be seen from the loss rate,so the effective prediction of the loss rate is of economic and social value.For the above situation,in thesis,LSTM neural network algorithm is used to build the driving behavior detection model,and the driving behavior detection system is designed and built to detect the driving behavior safety.The attentional BP neural network algorithm is used to build the prediction model of the loss ratio.The loss ratio prediction system is designed and built to predict the loss ratio.The main work has the following several aspects:(1)Design and construct driving behavior detection system based on LSTM neural network.The pulse sensor is used to obtain the driver’s pulse information,and the average sliding method is used to filter the noise of pulse data,then using the feature extraction which is based on the idea of statistical analysis to the processed data,extracting the main wave amplitude change rate,main wave drop rate and pulse rate change rate,constructing driving behavior detection model based on LSTM neural network,through the extracted characteristic parameters to conduct model training,designing and constructing the driving behavior detection system to detect and display the results of drivers’ driving behavior,finally,the success rate of the detection algorithm for the detection of angry driving behavior and normal driving behavior is compared and analyzed.(2)An attention-BP neural network based loss ratio prediction system was designed and constructed.First by collecting the compensation of insurance company status data sets,through using z-score processing data,the attentional mechanism is introduced into the traditional BP neural network to construct a prediction model based on attentional BP neural network,designing and constructing the loss ratio prediction management system,the system is used to predict the loss rate,and the accuracy of prediction is analyzed and summarized.(3)Thesis analyzes and summarizes the research status of driving behavior safety detection and prediction of loss rate,through the background research of road traffic development,his leads to the work of this paper-designing and constructing a system to carry out driving behavior safety test and prediction of loss rate.