基于伪谱法的城轨列车虚拟编队优化控制

Optimal Control for Virtual Coupling in Urban Rail Transit Based on the Pseudospectral Method

当前城市轨道交通网络面临巨大客运压力,虚拟编队技术的出现为提高城市轨道交通服务水平提供了一种解决方案。考虑虚拟编队列车具有速度趋向一致,间隔距离趋向最小的特征,本文提出了一种基于伪谱法的城市轨道交通列车虚拟编队优化控制策略,从而实现列车虚拟编队追踪运行曲线的实时生成与调整。首先,本文以最小化列车总运行时间为目标构建最优控制模型,离线计算得到列车运行参考速度曲线。然后,根据虚拟编队领导列车及跟随列车的不同追踪目标及约束条件,基于模型预测控制算法框架,分别建立领导列车和跟随列车的最优控制模型,并设计伪谱法进行高效实时求解计算。最后,本文以北京地铁昌平线为例,对比分析了基于虚拟编队技术与传统移动闭塞技术的列车优化控制策略。仿真结果表明,虚拟编队相对制动模式下的列车平均运行间隔距离及到站时间延迟显著小于移动闭塞绝对制动模式下的列车平均运行间隔距离及到站时间延迟,验证了本文所提出模型与方法的有效性。

The emergence of virtual coupling provides a solution to improve the level of service on urban rail networks,which are currently under enormous passenger pressure.Given that trains in a virtual coupling has the characteristics of consistent speed and minimum interval distance,in this paper we propose a distributed optimal control strategy for virtual coupling in urban rail transit based on the pseudospectral method according to time-varying tracking profiles.Thus,it is possible to achieve real-time generation and dynamic adjustment of tracking profiles for virtual coupling.First,to decrease the interval,minimization of the total train running time is set as the objective,and the reference speed profile is obtained by offline calculation.Then,according to the different objectives and constraints of the leading and following trains in a virtual coupling,optimal control models for the leading and following trains are established based on the algorithm architecture of model predictive control,and the pseudospectral method is designed to efficiently solve the sub-problem.Finally,we compare control strategies based on virtual coupling and moving blocking taking Beijing Changping Metro Line as an example.The simulation results show that the average interval distance and arrival time delay with the relative braking mode of virtual coupling are significantly smaller than those with the absolute braking mode of moving blocking.This confirms the effectiveness of the model and method presented in this paper.