基于遗传算法的发动机多元回归控制模型  

Multiple Regression Control Model Based on Genetic Arithmetic for Car-engines

作　　者：陈渝光[1] 李山[1] 廖仕利[1] 赵玮霖[2] 

机构地区：[1]重庆工学院电子信息与自动化学院 [2]重庆工学院材料科学与工程学院

出　　处：《农业机械学报》2006年第6期9-12,共4页Transactions of the Chinese Society for Agricultural Machinery

基　　金：重庆市科委应用基金资助项目(项目编号:7244);重庆市教委应用基金资助项目(项目编号:030606)

摘　　要：针对汽车发动机参数具有离散性、非线性和不确定性,采用常规的平均值控制模型难以进一步提高发动机动力性、燃油经济性等性能指标的问题,提出了基于遗传算法的发动机多元回归控制模型。该模型很好地融合了参数优化匹配在特定工况点处的优化功能和遗传算法优秀的回归参数寻优功能,形成了一种混合优化控制方法。探讨了基于遗传算法的回归参数自寻优,通过实验验证了该方法的可行性和可靠性,并表明动力性和经济性指标取得了明显改进。Aimed at the problem of car-engine parameters with discretion, nonlinear and uncertain characteristics, with which it is difficult to further improve the dynamics, economic benefits and exhaust performances by using general average control model, a new control strategy of the multiple regression control model based on genetic arithmetic was proposed. It well absorbed the advantages of the good parameters optimizing function for the given job status in parameters optimizing matching as well as the good regression parameters optimizing ability in GA, which created a new control method of mixed optimization. The regression parameters self-optimizing base on GA has been explored. Furthermore, the feasibility and reliability have been verified by the experimental results, it demonstrated that the dynamic indexes and the economic performance indexes are all improved obviously.

关 键 词：发动机 多元回归 遗传算法 自动寻优 

分 类 号：U464.11[机械工程—车辆工程] TP273.4[交通运输工程—载运工具运用工程]