摘要
通过建立定时约束条件下的最小能量控制模型,采用Pontryagain极小值原理推导了城市轨道列车节能操纵策略的组成。提出一种变长实矩阵编码的多种群遗传算法进行列车节能运行优化:采用多质点的列车牵引仿真器模拟列车运行;对列车运行控制序列采用变长实数矩阵编码;引入基于退火选择的变长算子以增强算法的全局搜索能力;适应值共享保持种群的多样性;多种群并行寻优提高收敛速度,增强寻优过程的稳定性。实例计算结果证实了该方法的有效性和先进性。
Through constructing the minimum energy consumption model with fixed running time,the urban railway energy saving train control strategies were obtained by Pontryagain minimum principle. An approach for solving train energy saving optimization problem based on variable-length real matrix coded multi-population genetic algorithm was proposed. A multi-particle train simulator was adopted to emulating train run. The GA chromosome consists of a variable two dimensional real number representing the train control sequence. A variable operator based on annealing selection was introduced to enhance global search performance. Fitness sharing keeps population＇s multiplicity. Multi-population parallel search improves convergence rate and evolution stability. An example proves the validity and advancement of the method.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第4期921-925,共5页
Journal of System Simulation
基金
铁道部重大课题(Z2006-04F)
国家电网公司重点课题(SGKJ[2007]102)
关键词
节能控制
多模态优化
矩阵实数编码
多种群
遗传算法
energy saving control
multi-model optimization
matrix real-coded
multi-population
genetic algorithm
