Route Prediction Based On Travel Habits And Real-time Traffic Conditions

With the development of social economy and the improvement of people's living standard,the number of urban residents is increasing.The problem of urban traffic is constantly highlighting.Efficient route guidance is one of the effective ways to improve the utilization ratio of road resources and the speed of traffic operation and relieve congestion.However,according to the research results at home and abroad,most of the existing route guidance systems are based on the dynamic route calculation only according to the real time road conditions,and do not consider the compliance and adjustment of the route by the traveller according to the subjective travel habits.In this paper,we will consider the real-time road conditions and the subjective habits of the travelers,predict the path selection behavior,establish the route selection behavior prediction model considering the road conditions and habits,and provide theoretical support for the real-time calculation of the induced route.This paper first conducts a multi day trip route RP survey.The results show that there is a strong correlation between the actual multi day route selection of residents' travel and personal travel habits.Based on this,the SP survey of habit and road condition compliance was carried out.The results showed that the dependence on habits and road conditions was different at different road conditions.Based on the multi day travel route RP survey data,this paper establishes a habit based Markov forecast model.The Markov method is used to predict the route selection probability of high frequency travel,such as going to work and going home at work days,shopping and going home at weekend and other high frequency trips.The result of model verification shows that the accuracy of working day work day and work day travel route forecast is low,68.27% and 65.38% respectively.The rate of travel route forecast for weekend shopping and weekend shopping is higher,81.73% and 87.50% respectively.Part of the reason may be that the way of commuting on weekdays is affected by morning and evening rush hour congestion.Based on the SP survey data of habit and road condition compliance,Logistic models were established to predict the compliance of travel section selection to habits and road conditions.The results of model validation show that the overall hit rate of the model is 81.40%.It shows that the model can predict the selection behavior of travelers at customary and non customary roads at different road conditions with high accuracy.Combining the route selection behavior prediction model based on the habit and the road selection behavior prediction model considering habits and road conditions,route selection behavior prediction process considering habits and road conditions is proposed at the end of the paper.The results of model verification showed that the rate of travel route forecast for going to work and going home at work days,shopping and going home at weekend was 82.69%,78.85%,86.54% and 89.42% respectively.It shows that based on customary route selection,travellers can improve the prediction accuracy of route selection based on real-time traffic adjustment to customary roads.The travel route prediction model established in this paper has a certain innovation compared with the traditional route prediction theory because it combines the travelers' compliance with the habits and the link adjustment according to the real time road conditions.The research results can be applied to the vehicle or mobile navigation APP.On the basis of the multi day route selection data,the customary route and selection probability are automatically generated by the program,and the route selection is dynamically adjusted according to the real-time road conditions,so as to quickly and accurately recommend the route and the peripheral facilities for the traveler.More importantly,because the model built in this paper better fit the route selection behavior of the traveler,it can provide a certain theoretical basis for improving the prediction accuracy of short time traffic flow under the large data background to a certain extent,and seeking the path guidance scheme which makes the traffic system with the highest efficiency.