基于多目标免疫算法的变刚度悬架的联合优化

Combined Optimization of Variable Stiffness Suspension Based on MISA

摘要:

为改善某装有变刚度悬架的轻型客车的平顺性和操纵稳定性，采用MATLAB编制MISA(multi objective immune system algorithm)的优化程序,与使用Adams Car建立的整车多体动力学模型组成联合优化模型，对该车前后悬架参数进行了优化.为保证模型的准确性,特对该车轮胎进行了力学性能测试，并通过参数辨识得到了基于魔术公式的轮胎属性文件(Pac2002).模型优化的抗体变量包括前悬架扭杆的扭转刚度、前后减振器的阻尼曲线系数、前后稳定杆的扭转刚度和后悬架钢板弹簧的两级刚度，并以该车实际情况给出了抗体的约束范围和便于编程优化的约束条件.优化目标抗原为国标下的稳态回转试验所达到的稳态最大侧向加速度、基准车速下的蛇行试验的车身横摆角速度和侧倾角，以及80 km·h-1、B级路面的前后悬架上方的车架大梁z向加速度均方根值.最后,根据优化结果试制了悬架样件，并在某汽车试验场进行了改进样车的平顺性试验、稳态回转试验和主观评价试验.试验结果表明，所采用的联合优化方法正确可行.这也反映了汽车计算机辅助优化(CAO)技术发展的一种趋势，对未来汽车底盘的虚拟开发及优化具有一定的指导作用.

Abstract:

In order to improve the ride comfort and stability of one light bus with two level variable stiffness rear suspension, a virtual combined optimal model was composed of the multi objective immune system algorithm（MISA) program composition by MATLAB and a virtual dynamic optimal model of the light bus in Adams Car. The parameters of the front and rear suspension were reasonably optimized. To ensure the accuracy of the model, the tire mechanical characteristics were tested, then the Magic Formula tire model was obtained by parameter identification method. An optimal method for the light bus’s suspension systems was put forward. In the proposed method, the torsion stiffness of front torsion bar, the front and rear damping curve coefficients, the torsional stiffness of front and rear stabilizer bars and the two level stiffness of rear suspension were taken as the optimal antibodies(variables). The max lateral acceleration of steady static circular simulation, the yaw rate and roll angle of the body of slalom simulation, and the Z direction acceleration root mean square(RMS) of the frame on B class road(80 km·h-1) were selected as the optimal antigen target. By using MISA to conduct optimization, the optimized suspension parameters were obtained. Finally, the suspension samples were manufactured and the comparative trials of vehicle ride comfort, handling and stability and subjective evaluation test were respectively carried out in an automotive proving ground to evaluate the optimization results. The results show that the proposed optimization method can be used for improving the vehicle’s ride comfort and stability and it has become a development trend of automotive computer aided optimization(CAO) technology. Therefore, it will provide guiding significance for the virtual development and optimization of automotive chassis.