采用最小信息损失方法的汽车主动悬架控制器降阶研究

A Study on the Order Reduction of Vehicle Active Suspension Controller by Using Minimum Information Loss Method

针对汽车悬架系统控制器阶数高、工程上难以实现的问题,建立了整车7自由度主动悬架模型,考虑人体对振动的敏感频段,设计了H∞加权控制器,以保证闭环系统稳定且具有较好的抗干扰性能。基于最小信息损失方法对20阶主动悬架控制器进行降阶研究。通过对降阶前后主动悬架闭环控制系统频域特性和乘坐舒适性的对比表明,将20阶主动悬架控制器降至8阶后,控制器状态的能控与能观信息损失之和小于50%,且悬架控制系统仍具有十分相近的控制性能。

In view of the problems of high order and difficulty in engineering implementation of vehicle suspension controller,a 7 DOF active suspension model is built,and with consideration of the sensitive frequency range of human body to vibration,the weighted H∞ controller is designed to ensure the good stability and disturbance rejection performance of closed-loop system.By using minimum information loss method,the order reduction for 20th order active suspension controller is studied.The results of comparison between active suspension closed-loop control systems before and after order reduction in terms of frequency characteristics and ride comfort demonstrate that with the controller order of active suspension system reduced from 20 to 8,the sum of controllable and observable state information loss is less than 50 % while the control performance of controller kept basically similar.