Research On Bus Stop Time Prediction With Or Without Lane

As an important part of the bus operation process,bus stop is an important facility in the bus operation line.The research and analysis of bus stop time can better grasp the operation law of the bus,improve the operation efficiency of the bus,and help optimize the configuration of the bus stop and the dynamic scheduling of the bus.Bus stop time is the main part of the whole bus operation time.Forecasting bus stop time is conducive to get the bus arrival time more accurately,which provides convenience for passengers,and plays a great role in improving the level of bus service system and bus operation management.Therefore,the research and analysis of bus stop time has practical and important significance.Firstly,this paper defines the bus stop with or without bus lane,studies and analyzes the process of bus stop,and divides the process of bus stop into three stages: accelerating exit,decelerating in and stopping service in the station.Then,the paper studies and analyzes the rules and characteristics of the stop time of the bus stops,mainly from the time period characteristics,cycle characteristics and line characteristics.This paper introduces the method of data acquisition and the method of bus stop time acquisition.Based on a large number of bus stop time data collected,this paper analyzes the bus stops with or without bus lanes,and studies the correlation of bus stop time at the stops.This paper introduces in detail from three aspects: different stations,the influence of the front car and the passenger flow,which provides support for the later model to predict the time of the stop.Finally,the idea of the exponential smoothing method is introduced.According to the above factors of bus stop time and the differences between two bus stops with or without bus lane,the influence factor model is established,and the exponential smoothing method model for predicting bus stop time at stops is constructed.Through this method,the predicted time of two different types of stations is obtained,and the predicted value and the actual value are analyzed;This paper introduces the BP neural network method,according to the analysis of the influencing factors,obtains four input layers: the time of the front parking station,the time of the historical parking station,the time of the front parking station and the number of people who swipe their cards,and establishes BP neural network method of genetic algorithm optimization prediction model.Then,the prediction value and the real value of the bus stoptime at the bus stop obtained by this method are analyzed in detail.Through the determination of the peak period,the evaluation index of the prediction results is established,the different prediction effects of different prediction models are compared and analyzed,and two better models of the prediction effects of different time periods and bus lanes or not are selected.After verification,the BP neural network method of genetic algorithm optimization prediction model has better prediction effect when the bus stop time data fluctuates greatly.The result of exponential smoothing method is better when the data of bus stop time is more stable.It can play a good role in promoting the research and analysis of bus stop time in the station.