模型跟踪广义预测鲁棒自适应控制器

本文采用滤波CARMA模型,基于内模原理,提出了一种新的广义预测鲁棒自适应控制器,并分析了闭环系统性能,在新的控制器中,引入适当的前馈作用,使得跟踪和调节问题解耦,利用部分状态跟踪、模型参考以及极点配置方法解决跟踪问题,利用多步预测滚动优化方法解决调节问题;适当选择滤波器可以保证对平稳随机扰动有满意的响应,减少可调参数对闭环系统响应的影响,增强系统对未建模动态的鲁棒性,仿真结果表明:该控制器对确定性和非平稳随机扰动具有不变性,对系统时延和阶次变化具有鲁棒性,适用于非最小相位和开环不稳定系统。

MODEL-FOLLOWING GENERALIZED PREDICTIVE CONTROLLER WITH ROBUSTNESS AND ADAPTIVENESS

It is presented in this paper a new. generalized predictive controller based on the internal model principle for the filtered CARMA model, a nice analysis of the closed-loop system resulting from this control law is made. The nwe controller introduces a proper feedforward term which enables the tracking and regulation performance capabilities to be treated separately. The tracking capability is obtained by partial state tracking, multiple model reference and pole placement approach, which can be suitably chosen by the user to achieve a different desired tracking dynamics; while the regulation performance is provided by the multistep receding horizon control approach which minimizes the tracking error. The closed-loop response can be improved by designing the filter which reduces the deleterious effects of unmodeled dynamics and disturbances and gives designer more freedom in choosing the design parameters. Simulations show that the controller has a strong robustness versus disturbance, variations of dead-time and model order. It is suitable for systems that are both unstable and/or in non-minimum phase.