基于切换和混合策略的多模型自适应控制

Multiple-models Adaptive Control with Mixing and Switching

混合与切换是多模型控制器的两种主要设计方法,混合控制不能及时响应参数突变,切换控制的暂态响应不理想。因此提出了一种新的多模型自适应控制结构。首先利用各模型子集的中心点值代替未知参数值得到最小误差,并与当前参数估计值得到的误差进行比较,判断参数是否发生跳变。若参数没有发生跳变,采用混合控制;若参数发生跳变,则把参数估计值重构为子集的中心点值,控制器就切换到对应的子控制器,及时响应参数变化。最后的仿真结果表明,在参数发生跳变时,该方法的控制效果较混合控制有明显的改善。

mixing and switching are two main design methods of multiple model controllers. The mixing control cant respond parame- tersstep timely, while the switching control makes the transient response undesired. Therefore a new multiple models adaptive control structure proposed in this paper. First, in order to judge whether the parameters have jumped, compare the error gained by current esti- mated values of parameter with the minimum errors obtained by replacing the unknown parameters with the values of each subset center. If there is no jump, the mixing control is adopted. Otherwise, replace the center value of subsets with the estimated values. As a result, the controller will switch to the corresponding candidate controller, responding to the jump timely. Finally, the simulation results show that when the jump happens, the method of the multiple models adaptive control with mixing and switching is much better than that of the mixing control.