Dispatching Optimization Of Urban Bus Based On Public Transport Data

Intelligent bus dispatching is of great significance to improve the efficiency of passenger travel and the level of bus service,and the data of real-time passenger flow is the basis of bus dispatch.This thesis relies on massive public transportation data provided by enterprises commissioned project " Operation Monitoring oriented Public Transport Big Data Analysis Technology and Application Research ",which research on the acquisition method of passenger flow and bus dispatch optimization.The identification of passengers' boarding and alighting stations is a critical step in the passenger flow acquisition.A bus passenger boarding site identification method based on spatio-temporal matching for flat fare lines is proposed,whose recognition accuracy rate reaches 98.7% by integrating GPS data and IC card data.Aiming at the problems caused by existing methods of identifying alighting site,the method based on the theory of travel chain may failed to recognize due to breakage of travel chain,methods based on site drop-off probability depends on subjective factors.A method based on historical trip characteristics is proposed in this thesis.The experimental results show that the accuracy of the method is higher and meets the application requirements.A short-term passenger flow forecasting method based on singular spectrum analysis and autoregressive moving average model is proposed according to the characteristics of short-term bus passenger flow that is both cyclical,trending,and highly random.The experimental results show the advantages of this method over traditional methods such as BP neural network and support vector machine in prediction accuracy and execution time.In the optimization model of bus scheduling,this thesis first determines the bus departure interval based on historic passenger flow,and then judging the abnormal passenger flow based on the short-term passenger flow forecast results.The optimization model of bus station scheduling is to minimize the cost of passengers and bus company.The concept of dominance is introduced in the solution algorithm to improve the elite strategy of NSGA-II and improve the uniformity of individual distribution in Pareto optimal solutions.And the feasible solution discriminant function is added to output the optimal compromise solution.Finally,a bus dispatching optimization model based on the ascending direction of the 346 bus in Chongqing is analyzed.The results show that the optimized plan can reduce passenger travel costs by approximately 6.6%.The current research results have been integrated into the cooperative's integrated traffic operation and monitoring platform.