多维并行遗传算法在列车追踪运行节能优化中的应用

Application of multi-dimension parallel genetic algorithm to energy-saving optimum control of trains in following operation

为了研究先行列车与追踪列车在移动闭塞信号系统控制下进行追踪运行时的综合节能优化控制问题,构建了以能耗与运行时间误差为目标的列车节能控制模型。该模型以列车的操纵手柄级位与列车的工况转换点(即操纵手柄级位改变时,列车的位置)为控制变量。在此基础上,提出了静态与动态速度约束条件(即线路的静态速度约束与先行列车的运行状态对其后的追踪列车所产生的动态速度约束),结合外部惩罚函数,利用多维并行遗传算法对该问题进行了求解。求解时,根据GA的进化代数动态的调整交叉与变异概率,提高粗粒度搜索与细粒度搜索的效率。同时,采用了坡道三分法和实数编码法来缩短染色体长度,提高收敛速度。最后,在移动闭塞信号系统列车运行仿真平台上验证了该优化控制算法的正确性与有效性。

In order to study the energy-saving optimum control strategy of a leading train and a tracing train in following operation under a moving block system,an energy-saving control model of trains is created.The aims of the model are energy consumption and trip time error.The control variables of this model are the operating handle level and the train’s position when the operating handle level is changed.Based on the model,the static and dynamic speed restraints are put forward.The static speed restraints are defined by the line conditions and the dynamic speed restraints of the tracing train caused by the leading train for the sake of safety.This problem is solved with the help of multi-dimension parallel genetic algorithm（GA） and external punishment function.During the solving process,the crossover probability and the mutation probability are adjusted dynamically according to the GA generation to improve the efficiency of the coarse grain search and the fine grain search.Ramps divided into three parts and the real number coding are adopted to shorten the length of chromosomes and improve the speed of convergence.Its correctness and effectiveness are validated at a simulation platform of train operation.