基于多混沌算子遗传算法的混合动力汽车控制策略优化  

Control Strategy Optimization for Hybrid Electric Vehicle Based on Multi-Chaotic Operators Genetic Algorithm

作　　者：梁俊毅[1] 张建龙[1] 马雪瑞[1] 殷承良[1] 

机构地区：[1]上海交通大学汽车电子控制技术国家工程实验室,上海200240

出　　处：《上海交通大学学报》2015年第4期442-449,456,共9页Journal of Shanghai Jiaotong University

摘　　要：为某款装备了电池/超级电容混合储能系统的并联型混合动力汽车设计了模糊控制策略.结合遗传算法的种群进化和混沌序列的随机遍历特性,将混沌初始化算子、混沌扰动算子、混沌局部搜索算子引入多目标非占优排序遗传算法(NSGA-II)中,构建了新的多混沌算子遗传算法(MCO-NSGA-II).运用MCO-NSGA-II算法进行了混合动力汽车模糊控制策略优化,以改进车辆的燃油经济性及HC、CO和NOx的排放性能.仿真结果表明,混沌初始化算子和混沌扰动算子可以改善原NSGA-II算法的搜索能力并增加种群多样性,而混沌局部搜索算子可以进一步增强算法局部搜索能力,能更好地搜索到理想的Pareto解集.运用MCO-NSGA-II算法进行优化,使混合动力汽车在欧洲城市驾驶循环(ECE)下的燃油消耗降低了11.8%,HC、CO和NOx排放分别下降了7.72%、15.72%和11.77%.This paper presented a fuzzy logic control strategy for a parallel HEV equipped with battery/ultra capacitor based hybrid energy storage system. By combining the population evolution feature of genetic algorithm and the randomicity and ergodicity of chaos sequence, the chaotic initialization, disturbance and local search operators were introduced into non-donminated sorting genetic algorithm-II（NSGA-II） to con- struct a novel multi-chaotic operators NSGA-II （MCO-NSGA-II）. MCO-NSGA-II was adopted to optimize the fuzzy control strategy for improving the fuel economy and emission performance of the target HEV. The results demonstrate that chaotic initialization, disturbance operators can improve the searching ability of NSGA-II and increase the diversity of the solutions. The chaotic local search operator can further im- prove the local searching ability to obtain better pareto solutions. By adopting MCO-NSGA-II, the fuel consumption of HEV under ECE driving cycle is reduced by 11.8% while the HC, CO and NOx emissions of HEV are decreased by 7.72~, 15.72~ and 11.77%.

关 键 词：混沌算子 遗传算法 多目标优化 混合动力汽车 

分 类 号：U469.72[机械工程—车辆工程]