摘要
动态矩阵控制(DMC)是模型预测控制的一种典型算法,它以被控对象的单位阶跃响应为预测和控制的模型,通过在线滚动优化和反馈校正来实施优化控制。对于存在状态等式约束的线性系统,文中在解决系统状态重构问题时改进了传统观测器,构造了投影卡尔曼滤波器,使系统的状态更好的符合约束条件,增强系统的稳定性。同时与概率约束算法进行比较,Matlab仿真结果证明了其优越性。
Dynamic matrix control （DMC） is a typical control algorithm of model predictive control, it uses the step response as model to predict and control. Through on-line rolling optimization and feedback rectification, the optimization of controlled system is implemented. This paper deals with the state estimation problem for linear system with linear state equality constrains by constructing an improved Kalrnan filter called projected Kalman filter, constructing the optimal estimate, which also satisfies linear equality constraints. The system stability is enhanced and the efficiency of this approach is demonstrated by comparing with an algorithm of probabilistic constrains into DMC. The simulation result also proves that.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2009年第14期95-98,共4页
Journal of Wuhan University of Technology
基金
南京信息工程大学重点科研基金(20070089)
