基于正交试验设计和多目标遗传算法的HEV参数优化  

Parameters optimization of hybrid electric vehicle based on orthogonal experimental design and multi-objective genetic algorithm

作　　者：周云山[1] 贾杰锋[1] 

机构地区：[1]湖南大学汽车电子与控制技术教育部工程研究中心

出　　处：《汽车安全与节能学报》2014年第4期324-330,共7页Journal of Automotive Safety and Energy

基　　金：国家"八六三"高技术研究发展计划(2012AA111710)

摘　　要：为在满足动力性前提下,降低混合动力汽车(HEV)的油耗和排放,提出了一种新的参数优化方法。以ADVISOR为仿真平台,应用正交试验设计,找出了对油耗和排放性能影响最显著的5个动力系统部件及控制策略的系统参数。建立了HEV多目标优化模型。用多目标遗传算法和最小二乘意义下的主客观组合赋权法,得到该模型的Pareto最优解集合,并从中选出了最优参数组合。结果表明:与优化前相比较,优化后的参数下,每100 km的油耗降低25.3%,每1 km的CO的排放质量降低35.5%,每1 km的HC+NOx的排放质量降低13.7%。因而,验证了该方法的有效性。A parameter optimization method for hybrid electric vehicle （HEV） was proposed to improve fuel economy and reduce emission within requisite power performances. An orthogonal experimental design was used with ADVISOR platform to ifnd out the ifrst iffth notable system parameters, which severely inlfuence the fuel economy and emission of HEV, among power components and control strategies. An optimization model was built using a multi-objective genetic algorithm to obtain a set of Pareto-optimal solution. An optimal parameter combination from the solution set was selected using a combination weighting method between subjective and objective evaluation in a least squares sense. The results show that with the optimized parameters, the fuel consumption per 100km is reduced by 25.3%, the CO emissions per kilometer is reduced by 35.5%, and the total HC and NOx emission is reduced by 13.7%. These facts verify the effectiveness of the method.

关 键 词：混合动力汽车 正交试验设计 多目标遗传算法 PARETO最优 参数优化 

分 类 号：U469.72[机械工程—车辆工程]