Series-parallel Hybrid Vehicle Control Strategy Design and Optimization Using Real-valued Genetic Algorithm

作     者：XIONG Weiwei YIN Chengliang ZHANG Yong ZHANG Jianlong 

作者机构：Institute of Automotive Engineering Shanghai Jiao Tong University Shanghai 200240 China 

出 版 物：《Chinese Journal of Mechanical Engineering》 (中国机械工程学报（英文版）)

年 卷 期：2009年第22卷第6期

页      面：862-868页

核心收录：

学科分类：08[工学] 0807[工学-动力工程及工程热物理] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

基　　金：supported by National Hi-tech Research and Development Program of China (863 Program  Grant No. 2006AA11A127) 

主　　题：series-parallel hybrid electric vehicle control strategy design optimization real-valued genetic algorithm 

摘      要：Despite the series-parallel hybrid electric vehicle inherits the performance advantages from both series and parallel hybrid electric vehicle, few researches about the series-parallel hybrid electric vehicle have been revealed because of its complex co nstruction and control strategy. In this paper, a series-parallel hybrid electric bus as well as its control strategy is revealed, and a control parameter optimization approach using the real-valued genetic algorithm is proposed. The optimization objective is to minimize the fuel consumption while sustain the battery state of charge, a tangent penalty function of state of charge（SOC） is embodied in the objective function to recast this multi-objective nonlinear optimization problem as a single linear optimization problem. For this strategy, the vehicle operating mode is switched based on the vehicle speed, and an ＂optimal line＂ typed strategy is designed for the parallel control. The optimization parameters include the speed threshold for mode switching, the highest state of charge allowed, the lowest state of charge allowed and the scale factor of the engine optimal torque to the engine maximum torque at a rotational speed. They are optimized through numerical experiments based on real-value genes, arithmetic crossover and mutation operators. The hybrid bus has been evaluated at the Chinese Transit Bus City Driving Cycle via road test, in which a control area network-based monitor system was used to trace the driving schedule. The test result shows that this approach is feasible for the control parameter optimization. This approach can be applied to not only the novel construction presented in this paper, but also other types of hybrid electric vehicles.