基于响应面和粒子群算法的悬置多目标优化  

Multi-objective Optimization of Cab Suspension Based on Response Surface and PSO

作　　者：关挺 钟绍华[1,2] 周才[1,2] 范文涛[1,2] 

机构地区：[1]武汉理工大学现代汽车零部件技术湖北省重点实验室,湖北武汉430070 [2]武汉理工大学汽车零部件技术湖北省协同创新中心,湖北武汉430070

出　　处：《物流技术》2017年第10期101-108,共8页Logistics Technology

摘　　要：为了改善某厂商驾驶室悬置平顺性较差的问题,进行了实车道路平顺性实验并选出了平顺性较好的结构,并提出了一种新的优化方法优化了悬置的参数:在ADAMS中建立了参数化模型,以驾驶室地板垂向加权加速度均方值和质心俯仰加权加速度均方值作为响应评价指标,用Box-Behnken Design(BBD)试验设计方法进行了多次仿真,根据仿真结果得到了两个响应量的响应面,并根据最小二乘法拟合出了相应的响应多项式方程,对两个响应量取不同权重,用粒子群算法(PSO)对目标函数进行了优化。结果表明,优化后的垂向加权加速度均方值和俯仰加权加速度均方值有明显下降,系统平顺性得到提高。In this paper, to improve the ride comfort of the cab suspension of a certain carmaker, we carried out a vehicle road test to identify the suspension structure with better ride comfort and proposed a new method to optimize the parameters of the suspension. More specifically, we established the parameterlzed model in ADAMS, then made the mean square value of the weighted acceleration of the floor vertical and centroid pitching directions of the cab floor as the response evaluation indexes, and performed multiple simulations using the Box-Behnken Design test methodology, through which, the response surface of the two response variables were obtained, the corresponding polynomial equation derived and the objective functions optimized. Through the simulation, we demonstrated obvious drop in both values after the optimization.

关 键 词：驾驶室悬置 平顺性试验 响应面 最小二乘法 粒子群算法 BBD试验 

分 类 号：F224.0[经济管理—国民经济] F253.9