Multi-objective Optimization Of Timetable For Urban Rail Transit Based On The Stochastic Passenger Flow

With the increase of urban population and private cars in China,the traffic problem and the air quality are getting worse.Therefore,the demand for public transport is growing.Owing to the characteristics of high efficiency,large capacity and convenience,the urban rail transit has become the most important passenger transport mode.However,the random distribution characteristics of passenger flow often leads to unreasonable matching between passenger flow and the train flow,which reduces the efficiency of the train operation and the satisfaction of passengers.How to develop the train schedule to adapt to the random passenger flow and achieve the collaborative optimization between operating costs and passenger satisfaction are the urgent problems to be solved in the current urban rail transit.Train traction energy consumption is an important factor in determining the operating costs of urban rail transit,so it should also be considered in the timetable optimization to improve the overall performance of the train operation plan.Optimizing the train schedule scientifically and reasonably plays a significant role in improving the overall performance of the schedule.Through analyzing of the distribution characteristics of the stochastic flow,the timetable is optimized for multi-objective in this paper.Firstly,the model of multi-objective optimization of timetable based on stochastic flow is formulated and solved.The passenger expectation and variance OD(Original-Destination)matrices are set up through analyzing the passenger data for Beijing subway lines of Automatic Fare Collection system.The normal distribution of passenger flow is verified with the method of test goodness for fit.Taking the running time,departure interval,dwell time as decision variables,the model of multi-objective optimization of timetable based on stochastic flow is established.The optimization objectives include:the waiting time for passenger,the difference between passenger on load and the desired value,the train running balance and the robustness of timetable.An improved genetic algorithm is used to solve the model,and a timetable with better comprehensive performance can be obtained.Secondly,based on the timetable optimization model of the stochastic passenger flow,the two-stage energy saving optimization method is carried out on the timetable.Keeping the inter-station running time of the optimized schedule unchanged,the trajectory is optimized adopting a coasting optimization algorithm.Then considering the use of regenerative braking energy through the train coordination,a new optimization model is built and solved by improved genetic algorithm,in which the synchronization time for bidirectional trains,the dwell time,the departure interval are defined as the decision variables.The utilization of regenerative braking energy is one of the optimization objectives.Finally,the simulation module of hardware-in-the-loop for train timetable is developed,which integrates multi-objective optimization of timetable and the coasting optimization algorithm.The main functions of the module include:timetable optimization,train simulation,communication with the simulation driving platform and evaluation of timetable.Through the transmission of relevant data between modules,the simulations of the trains running in typical scenarios can be completed.