一种用于主动悬架LQG控制器设计权重优化的改进遗传算法  

A modified genetic algorithm of weighting optimization in LQG controller design for active suspensions

作　　者：段源博 李靖玮 罗建南 DUAN Yuanbo;LI Jingwei;LUO Jiannan(School of Mechatronic Engineering and Automation,Shanghai University,Shanghai 200444,China)

机构地区：[1]上海大学机电工程与自动化学院,上海200444

出　　处：《振动与冲击》2023年第11期278-283,共6页Journal of Vibration and Shock

摘　　要：如何实现在不同行驶条件下汽车两大性能的最佳协调(即乘坐舒适性和操作稳定性)是主动悬架设计的关键。在LQG控制器设计中,性能指标的加权系数决定了汽车在当前条件下的各性能间的最佳平衡。而实际中需要考虑的因素来自多个方面,如不同路面的不平度、车辆行驶速度、车载质量、甚至还可能包括不同驾驶员偏好等,因而使得控制器设计中对权重系数的选取较难处理。针对这一难题,提出了一种改进的遗传算法优化方案。根据驾驶人(或乘员)对不同方面的性能需求,建立包含不同权值系数的适应度函数,针对适应度函数设计了典型的优化模式,并通过对惩罚函数值的适当选取来实现对性能指标加权系数的优化。通过不同工况及要求下的仿真,分析了该算法的可行性和有效性。研究结果表明,所提出的权值适应度函数优化方法简单、可行、有效。所设计的改进遗传算法能够高效地优化性能指标的加权系数以实现汽车在当前条件下的不同性能的最佳平衡,从而为主动悬架LQG控制器的设计提供一个有效的方法。It is the key to obtain a best compromise for two important performances of automobile(i.e.,ride comfort and handling stability)in different operation conditions for designing active suspensions.In LQG controller design,the weighting coefficient of the performance index determines the best balance among the performances of the vehicle.In practice,many influence factors have to be considered,including road roughness,vehicle speed,loading condition,and even driver’s preference,etc.Hence,it is crucially important to properly select weighting values in LQG controller design.Aiming at the difficulty in determining the weighting parameters,a modified genetic-algorithm optimization method was proposed based on different performance requirements.The fitness function including different weighting values was built and a few of typical optimization schemes of fitness function were designed accordingly.By modifying penalty function,efficient weighting optimization was realized.Based on the established vehicle suspension model along with properly designed LQG controller,simulations were carried out for examining the feasibility and effectiveness of the algorithm.The results show that the optimization algorithm is feasible to obtain the best compromise among different performance requirements.The proposed method of weighted fitness function can provide an efficient way to select targeted weightings for LQG controller design.

关 键 词：主动悬架 LQG控制器设计 加权系数优化 改进遗传算法 

分 类 号：U461[机械工程—车辆工程]