非线性系统模型预测控制若干基本特点与主题回顾

Review of fundamental properties and topics of model predictive control for nonlinear systems

随着线性系统模型预测控制的大量成功应用,基于非线性模型的预测控制简称非线性预测控制(NMPC)在近20年来引起了广泛关注,并取得了丰富的研究成果.本文在阐述NMPC基本原理和特点的基础上,分别从优化可行性、稳定性、鲁棒性、优化求解、吸引域等主题出发,分析并总结NMPC的研究现状.最后,在指出NMPC有待进一步深入研究问题的同时,展望当前NMPC的一些研究方向.

With the successful applications of model predictive control for linear systems, predictive control for nonlin- ear models （NMPC） has received wide attention and achieved rich results in the last two decades. Based on the fundamental principle and characteristics of NMPC, the state of the art of NMPC is reviewed from the topics of feasibility, stability, ro- bustness, optimization solution to regions of attraction, respectively. Open problems concerning NMPC are also discussed.