遗传算法在混合动力汽车控制策略优化中的应用  

Application of Genetic Algorithm in Optimization of Control Strategy for Hybrid Electric Vehicles

作　　者：浦金欢[1] 殷承良[1] 张建武[1] 

机构地区：[1]上海交通大学,上海200030

出　　处：《中国机械工程》2005年第7期648-652,共5页China Mechanical Engineering

基　　金：国家 863 高技术研究发展计划资助项目(2001AA501200;2003AA501200)

摘　　要：提出了一种基于浮点数编码遗传算法的混合动力汽车控制策略参数优化新方法。以一辆实际混合动力汽车样车的逻辑门限控制策略为例,分析并建立了控制策略参数优化的有约束非线性规划模型,其目标函数包含最小化油耗和排放。提出了采用稳态进化模型和浮点数编码遗传算法的参数优化方法,用于求解一组最优的控制策略参数。仿真结果表明,该方法可以找到一组全局最优的参数,将其用作离线参数优化,可以大大缩短控制器的实车标定时间。A new method based on real-coded genetic algorithm (RCGA) was presented for parametric optimization of control strategy for hybrid electric vehicles. A logic threshold control strategy (LTCS) implemented in an existing hybrid car was considered. A set of parameters was adopted in the LTCS to command the operation of the vehicle. The optimal tuning of these parameters was formulated as a constrained nonlinear programming problem with an objective function minimizing both fuel consumption and emissions. A genetic algorithm using steady-state evolution model and real value coding was proposed and applied to find the optimal set of parameters. Experimental results demonstrate effectiveness of the proposed method, which can be used for off-line parametric optimization and thus shorten the time of controller calibration in the real vehicle.

关 键 词：混合动力汽车 控制策略 遗传算法 浮点数编码遗传算法 参数优化 

分 类 号：U469.72[机械工程—车辆工程]