基于多参数规划的显式模型预测系统设计的可行域扩张算法  

A multi-parametric-programming-based feasible region-expansion algorithm for synthesis of explicit model predictive control system

作　　者：张聚[1] 郝平[1] 王万良[1] 

机构地区：[1]浙江工业大学自动化研究所,浙江杭州310014

出　　处：《控制理论与应用》2009年第3期305-308,共4页Control Theory & Applications

基　　金：国家自然科学基金资助项目(60604014).

摘　　要：基于约束线性优化控制问题的多参数二次规划求解方法,提出设计显式模型预测控制系统的可行域逐步扩张算法.首先建立一种求取优化控制问题输出不变集的迭代算法.以该输出不变集作为多参数规划问题中状态区域约束限制的初始条件,通过反复求解多参数规划问题和不断改变状态区域约束限制,能够逐步扩大显式模型预测控制系统的无限时间可行区域,直到可行域不再继续扩大.算法收敛时设计得到的显式模型预测控制系统在其所有的状态分区上都是无限时间可行的.通过数值仿真计算,验证本文提出算法的有效性.Based on the multi-parametric quadratic programming（MPQP） for the linear constrained optimal control problem, a feasible and a feasible-region-expansion algorithm is proposed for synthesis of the explicit model predictive control（MPC） systems. Firstly, an iterative algorithm for computing the output invariant set of the optimal control problem is established; the computed invariant set is then used as the initial constraint for the state region in the MPQP problem. By iteratively solving the MPQP problem and replacing its constraint for the state region, the feasible region of the explicit MPC system is progressively expanded until its convergence. The state trajectory of the resulting explicit MPC systems starting from any state in its state partitions is always infinite time-feasible. Simulations also show that the proposed algorithm is effective.

关 键 词：多参数二次规划 显式模型预测控制系统 可行域逐步扩张法 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]