Energy-Efficient Multi-Train Control In Metro Transit System

Metro Transit system is featured by larger passenger capacity, convenience, safety, punctuality, comfort, cleanness and fewer land acquisition. With China’s rapid economic growth and continual accelerating urbanization, this system is believed to play a key role in relieving urban traffic congestion and promoting harmonized development of economics and society. However, Urban rail transit system still has high operating costs due to huge power consumption, where the traction power cost is about 2.4 kW·h/vehicle kilometers, accounting for 40% to 50% of total operating energy, which means that it is an essential issue for operator to reduce energy consumption, particularly the train traction energy. In urban rail transit system, traffic conflicts and energy exchange often exist among different trains, since it is a complex multi-train, multi-target real-time system. In this paper, the objective is to minimize the total energy consumption at power substation. Speed profile optimization for single train and multi-train are studied to develop energy-efficient train control algorithm by reusing regenerative energy. Different control strategies are tested to investigate the energy saving effects and application extent. The algorithms have finally been applied to the design of energy-efficient timetable. Therefore, this research has important academic value and practical significance.To begin with, based on theories of train traction calculation, automatic train control and traction power supply, energy-efficient multi-train control model to minimize total energy consumption at substations has been established by considering the influences of train traction/braking performance, operating mode, route parameters, signaling system, power supply system and the regenerative braking energy.Secondly, according to the single train mathematical model, an on-time and energy saving algorithm based on maximum principle and dynamic programming has been proposed to solve the single train optimal control problem, and has been verified through simulation. Single train optimal control has been analyzed theoretically:According to necessary conditions given by maximum principle and the characteristics of rail transit system, control principles and evaluation criterion are discussed thereby. In order to solve the problem through dynamic programming, the journey is converted to a multi-stage decision process. A real-time adjustment strategy for train operation is also proposed to improve the robustness of the algorithm.Next, in addition to minimizing traction energy for single train, the minimization of total energy consumption at substations is set as new objective when the regenerative braking energy is considered in two trains’ operating environment. If the late-departed train adjusts its speed according to the position, speed and regeneration potential of the early-departed train, more regenerative energy may be reused and total energy consumption can be reduced. To find the optimal control for the late-departed train, maximum principle is applied to study necessary conditions and control principle. Then, based on multi-train system and control model, dynamic programming has been used to search energy efficient control sequences for the late-departed train. The optimization effect is also analyzed and compared with the scenario that only the early-departed train is adjusted.Based on previous study, simultaneous optimization of speed profiles for both trains running in opposite directions is studied. With the dual objectives of maintaining schedule requirements and optimizing energy efficiency, dynamic and electric performance of two opposing trains operated in the same DC power section are discussed. Setting tractive/braking efforts of both trains as control variables, a multi-train coordinated control algorithm based on improved genetic algorithms has been developed to search for the optimal train speed profiles, due to the non-linear constraints and interdependence of variables. Simulation through Visual C++platform demonstrates that the algorithm can provide optimal train speed profiles with better energy performance while satisfying operational constraints. Compared with the scenario that only optimizes the late-departed train, coordinated train control helps to further reuse regenerative braking energy, but the improvement is limited.Finally, the multi-train control research has been applied to the design of energy-efficient timetable. Two train operating modes, one based on traction energy saving only; the other included the reuse regenerative energy, are implemented into the mixed integer programming for timetable design respectively. Simulation results show that for existing timetable, the application of regenerative energy based control algorithm helps to reduce the total energy of the system; while for energy-efficient timetable, this algorithm helps to reduce the impact of timetable adjustment to train operation.