Energy-efficient Train Schedule Optimization Based On Operation Levels

Urban rail transit is a safe,high-speed and low-carbon public transportation mode,the average energy consumption of which is smaller when compared with other transportation modes due to its large traffic volume.However,with the growth of population and the construction of the urban rail transit lines,the total energy consumption of urban rail transit increases gradually,which means that it is necessary to conduct energy-efficient optimization for urban rail transit.The traction energy consumption of a train running in a section between two stations is decided by the running time and the operation of the train in this section.With the aim of decreasing the traction energy consumption,on the one hand,a part of train services can be assigned to long running times in the train schedule,and on the other hand,the operation of a train in a section with a given running time can be optimized.In the current metro practice,the running times depend on the running levels stored in the Automatic Train Supervision(ATS)system,where each running level corresponds to a running time and a pre-stored reference speed trajectory in Automatic Train Operation(ATO)system.The ATO system drives the train automatically to ensure that its realized running time is equal to the running time of the running level assigned by the ATS system.In this thesis,the energyefficient train speed trajectory optimization problem is first investigated.A robust methodology is then proposed to optimize the running levels based on the stochastic operation scenarios.Based on the given running levels,the train schedule and rolling stock circulation plan are optimized simultaneously,without and with the consideration of regenerative energy utilization,to satisfy the passenger demand and to minimize the energy consumption.The main research topics in this thesis are listed as follows:(1)Train speed trajectory optimization based on the speed-distance network.According to the train dynamics,a discrete speed-distance network model for train speed trajectory optimization is presented by discretizing the speed and distance.By applying Lagrange relaxation,the original problem is transformed into a shortest path problem,which is solved by the dynamic programming.Based on the characteristics of train speed trajectory,the network reconstruction method is presented to reduce the required computation time for small discrete intervals.The computational results based on the data of Beijing Yizhuang metro line demonstrate the effectiveness of Lagrangian relaxation method and network reconstruction method when compared with other approaches.(2)Robust optimization of running levels based on stochastic operation scenarios.A mixed integer nonlinear programming model is proposed to optimize the running levels,with the objective of minimizing the energy consumption and the deviation from the planned train schedule caused by the dwell time disturbances.The mixed integer nonlinear programming model is transformed into a mixed integer linear model by introducing auxiliary variables and constraints.The optimization software CPLEX is then adopted to solve the resulted problem.Based on the passenger flow data of Beijing Subway Yizhuang line,the dwell time disturbance scenarios are generated stochastically,and multiple scenarios are involved in robustness optimization process.Based on the simulation results of train operations,the train rescheduling with optimized running levels can reduce the deviation more effectively than that with running levels with fixed time intervals.(3)Energy-efficient integrated train scheduling and rolling stock circulation planning.A mixed integer linear programming model is presented to optimize the train schedule and rolling stock circulation plan simultaneously,where the running level selection of train services are optimized to save energy and to satisfy the passenger demand,which is considered indirectly via the service pattern.To enhance the regularity of train schedule,two model extensions are proposed to ensure that train services are assigned to the same running levels during off-peak hours.Based on the data of Beijing Yizhuang line,numerical experiments are conducted for different train scheduling models.It can be seen from the results that the integrated train scheduling and rolling stock circulation planning model with the flexible running level selection achieves the best performance in terms of the objective function among all the models.Based on the energy-efficient train scheduling model,an extended model is proposed to take the utilization of regenerative braking energy into consideration,which is solved by a twostep method.According to the numerical experiments based on the data of Beijing Yizhuang line,compared with the train scheduling model without the consideration of regenerative energy utilization,the train scheduling model with regenerative energy utilization can improve the amount of regenerative energy utilization effectively.When the headway between train services is smaller,the regenerative energy utilization increases.In this paper,there are 39 figures and 16 tables.65 references are cited.