基于遗传算法的厢式货车平顺性优化  

Ride Optimization of Van Truck Based on Genetic Algorithm

作　　者：陆建辉[1] 周孔亢[1] 侯永涛[2] 于凯 嵇佳琪 

机构地区：[1]江苏大学汽车与交通工程学院,镇江212013 [2]江苏大学机械工程学院,镇江212013

出　　处：《机械工程学报》2017年第20期121-130,共10页Journal of Mechanical Engineering

基　　金：江苏省博士后基金(1101114C);江苏大学高级专业人才科研启动基金(10JDG064)资助项目

摘　　要：在Recur Dyn中建立某厢式货车的整车多体动力学模型。基于离散梁法,建立整车模型中的钢板弹簧动力学模型,在板簧模型中,采用用户自定义子程序的方法为弹簧片间定义的三向力,可有效模拟簧片间的接触力和摩擦力,应用基于元模型的优化方法,对钢板弹簧模型的参数进行辨识;基于AR模型开发的路面不平度重构程序,可生成能够被Recur Dyn调用的路面文件;进行该厢式货车整车平顺性试验,仿真和实车试验结果验证了所建整车模型的正确性;采用试验设计方法分析了悬架参数及驾驶室悬置参数对平顺性的影响,为平顺性优化时变量范围的选取提供了依据;提出一种采用批处理方式,通过Process Net集成AForge.Genetic组件实现遗传算法进行整车平顺性优化的方法,优化后驾驶员脚部地板处总加权加速度均方根值下降了9.35%,提高了厢式货车的平顺性能。The multi-body vehicle dynamics model of a is truck is built in Recur Dyn. By using discrete beam method, the multi-body dynamics model of leaf spring in vehicle model is established. The contact and friction force between the interacting spring leafs has been effectively simulates by defining the user subroutine translational forces. Based on meta-model optimization algorithm, the parameters of the two leaf spring dynamics model can be effectively identified. Based on AR model, a road roughness reconstruction application program is developed which can output the road file to be used in Recur Dyn. The ride comfort of the van transporter physical prototype is tested. Data of the simulation and the experiment are analyzed to verify correctness of full vehicle model. DOE method is used to analyze the influence of suspension parameters and cab suspension parameters on the ride comfort, which provides the basis for the selection of variable scope. A method to optimize the ride comfort of vehicle model is proposed. In this method, Genetic algorithm is realized by integrating the AForge.Genetic component with Process Net and using batch mode. After optimization, the total weighted RMS acceleration value which tested on the floor of driver’s foot decreased by 9.35% and the ride comfort of the van is improved.

关 键 词：厢式货车 平顺性 参数辨识 遗传算法 多体动力学 

分 类 号：U463[机械工程—车辆工程]