Study On Scheduling Strategy Of Changsha Metro Line 4

With the development of the society,the urban population and the traffic pressure keep increasing.The subway has gradually become one of the effective ways to solve the traffic pressure.It has many advantages,such as speed,punctuality,safety and low pollution.The network of subway is more and more complex,passenger flow of the subway will also have great changes.Therefore,it is necessary to change the scheduling strategy according to passenger flow.It is appropriate to decrease the operation interval of the subway during the rush hour and increase the operation interval during the flat hour.Try to reduce the company’s operating costs as the same as ensure the passenger satisfaction.This paper mainly studies the prediction of passenger flow and the optimization of train scheduling scheme.Firstly,the passenger flow characteristics of Changsha Metro Line 4 are analyzed based on two aspects with time and space.In terms of time,the passenger flow characteristics are analyzed in different periods of a day and a week.In terms of space,the passenger flow characteristics are analyzed in upstream,downstream and different sections.The unbalanced coefficient of passenger flow is calculated.This analysis provide a basis for passenger flow prediction and scheduling.Secondly,based on the historical short-term passenger flow data,the passenger flow of Changsha Metro Line 4 is predicted.Selecting BP neural network and Support vector machine method to predict passenger flow after analyzing,the most important factors affecting passenger flow(weather,holiday,day time,week)are selected as independent variables,cross section passenger flow as dependent variables,and relevant prediction models are established.Then normalize the data and input it into the prediction models to get passenger flow.By analyzing and comparing the accuracy of the two prediction methods,the prediction accuracy of support vector machine is high,and it is determined that the support vector machine is used to predict the passenger flow,the time-sharing passenger flow of each section of Changsha Metro Line 4 on July 31,2019 and the passenger flow in rush hour of a week are predicted.Finally,the train scheduling strategy is optimized from the time and space.In terms of time,the multi-objective optimization model is established considering the passenger experience and the cost of the subway company.The passenger waiting time,the number of trains and the full load rate are the constraints,Genetic algorithm is used to solve the scheduling model and determine the optimal travel interval finally.In terms of space,the section difference rate of each section is calculated,the appropriate intersection type and turnback station are determined combining the result with the crossing problem of Changsha Metro Line 4.By comparing the number of available vehicles,the size intersection mode can save costs better than the single mode.According to the optimization results in both time and space,the final optimization plan is developed.Through comparison with the original operation plan,the total mileage reduced 15.02% than before,and the number of train trips reduced 14.38%.The full load rate is improved while ensuring passenger comfort.The scheduling scheme provides theoretical support and reference suggestions for the scheduling strategy researchers of the metro company.