改进麻雀算法在列车ATO多目标优化中的应用  

Application of Improved Sparrow Algorithm in Multi-objective Optimization of Train ATO

作　　者：王一栋 肖宝弟 岳丽丽[1] 李茂青[1] 林俊亭[1] WANG Yidong;XIAO Baodi;YUE Lili;LI Maoqing;LIN Junting(School of Automation and Electrical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China;Beijing Consen Traffic Technology Co.,Ltd.,Beijing 101318,China)

机构地区：[1]兰州交通大学自动化与电气工程学院,兰州730070 [2]北京康吉森交通技术有限公司,北京101318

出　　处：《铁道标准设计》2024年第7期192-199,共8页Railway Standard Design

基　　金：中国国家铁路集团有限公司科技研究开发计划项目(N2021G045);甘肃省教育厅优秀研究生“创新之星”项目(2022CXZX-535)。

摘　　要：针对列车自动驾驶(ATO)运行过程多目标优化问题,以列车运行安全性、列车动力学模型等因素为约束条件,考虑列车准时性、能耗、舒适性等指标,使用模糊隶属度法建立多目标优化模型。利用罚函数处理约束条件,将停车误差与限速作为惩罚项并构造出适当的惩罚函数加入到目标函数中,从而得到增广目标函数,提出基于改进麻雀算法(ISSA)的求解策略。为提高麻雀算法(SSA)的全局寻优能力,避免收敛于局部最优,引入Logistic映射、自适应超参数、变异算子对传统麻雀算法进行改进,通过测试函数对算法性能进行验证,表明ISSA算法的收敛速度、寻优精度比传统SSA算法好。以工况转换点为决策变量,通过ISSA算法对工况转换点的位置及速度进行寻优,进而获得目标速度-距离曲线。最后选取城轨车辆参数与线路数据进行仿真验证,仿真结果表明:所提优化策略相较于未优化前,舒适性提高了21.22%,能耗降低了22.41%,准时性与停车误差满足要求;与PSO优化方法相比,收敛速度更快,运行时间几乎一样的情况下能耗降低了12.74%,节能效果更佳;停车误差降低了20.45%,舒适性保持在舒适范围之内;对于速度-距离曲线,巡航距离更长、惰行距离变短、最高运行速度降低。由此可见,达到了综合优化ATO的目的,验证了ISSA优化策略的有效性。Aiming at the multi-objective optimization problem of automatic train operation(ATO)process,a multi-objective optimization model is established by using fuzzy membership functions,taking train operation safety,train dynamics model and other factors as constraints,and considering train punctuality,energy consumption,comfort and other indicators.The penalty function is used to deal with the constraints.The parking error and speed limit are used as penalty terms and appropriate penalty functions are constructed and added to the original objective function to obtain the augmented objective function,and a solution strategy based on improved sparrow algorithm(ISSA)is proposed.In order to improve the global optimization ability of sparrow algorithm(SSA)and avoid converging to local optimum,Logistic mapping,adaptive hyper-parameters and mutation operator are introduced to improve the traditional sparrow algorithm.The performance of sparrow algorithm is verified by test functions,which shows that the convergence speed and optimization precision of ISSA algorithm are better than those of traditional SSA algorithm.Taking the operation mode transition point as the decision variable,the ISSA algorithm is used to optimize the position and speed of the operation mode transition point,and then the target speed distance curve is obtained.Finally,vehicle parameters and line data of urban rail transit are selected for simulation verification.The simulation results show that the comfort of the proposed optimization strategy is improved by 21.22%and the energy consumption is reduced by 22.41%compared with those before optimization.The punctuality and parking error meet the requirements.Compared with the PSO optimization method,the convergence speed is faster,the energy consumption is reduced by 12.74%under the condition of almost the same running time,and the energy saving effect is better.The parking error is reduced by 20.45%,and the comfort is maintained within the comfort range.For the speeddistance curve,the cruising distance is

关 键 词：城市轨道交通 列车自动驾驶 多目标优化 目标速度曲线 改进麻雀算法 模糊隶属度 罚函数 

分 类 号：U284.48[交通运输工程—交通信息工程及控制]