基于遗传算法的增程式电动车模糊控制器设计

摘要

A Multiple Input Single Output(MISO) fuzzy logic controller is designed for the energy distribution of a fuel cell Extended-Range Electric Vehicle(E-REV) on basis of Genetic Algorithm(GA). By using battery SoC and transient power demand at the load bus as inputs, the fuzzy controller can calculate the optimal output power of the fuel cell extended range, which promotes a rational distribution of energy resources. Genetic algorithm optimizes the fuzzy membership function and rules of the controller, so the human knowledge or experience for parameter settings of the controller can be avoided. The validation is achieved by the simulation of ADVISOR and the experiment of car roller bench. The results demonstrate that the proposed fuzzy controller can improve the energy management strategy to attain a better economic performance.%针对燃料电池增程式电动汽车动力系统双能量源间的分配问题，设计基于遗传算法的多输入单输出(MISO)模糊控制器。控制器将动力蓄电池SOC和负载总线实时需求功率作为输入变量，求解燃料电池增程器的最佳输出功率，从而获得蓄电池和燃料电池的输出功率分配关系，以此实现不同功率需求下车载多能量源间的合理分配。为克服传统模糊控制器的参数设置仅依靠专家经验设定的局限性，采用遗传算法对模糊控制器的隶属函数和控制规则参数进行优化设计。通过ADVISOR软件仿真和转鼓实验台实车验证，结果证明，与传统能量控制策略相比，优化设计后的模糊控制能量管理策略能够明显提高增程式电动汽车的燃料经济性，并表现出较好的工况适应能力。