摘要
二次型目标函数优化可得到状态反馈的最优控制律,实现系统的闭环最优控制,因此广泛应用于现代优化控制中。二次型目标函数的权矩阵决定最优控制律,但是由于权矩阵和系统状态方程的联系关系不明确,造成加权矩阵的选择过程复杂。采用系统基本状态方程推导二次型目标函数加权矩阵的代数表达式,由线性矩阵不等式工具在稳定裕度区域寻求系统控制律的迭代求解,可以方便的得到状态加权矩阵,并由此求得系统的最优控制律。仿真给出系统的控制效果比较,验证了算法的有效性,为二次型目标函数算法优化提供了一种可行的方法。
The quadratic objective function optimization can get the state feedback optimal control law, to realize the optimal control of closed - loop system, therefore it is used in modem optimal control widely. The weight matrix of quadratic objective function can determine the optimal control law, but because the relation between the weight matrix and the system state equation is not clear, the weighting matrix choice process is complex. The basic state equation was used to derive the weighted matrix algebraic expressions of quadratic objective function, and the iterative solutions of the system control law was solved in o stability margin area with linear matrix inequality（lmi） tools to determine the weighting matrix easily, so the system optimal control law can be obtained. Simulations gave the compared results of the system control effects, which verifies the effectiveness of the algorithm.
作者
邢丽娟
杨世忠
XING Li - juan;YANG Shi - zhong(Automation Engineering College, Qingdao Technological University, Qingdao Shandong 266520, Chin)
出处
《计算机仿真》
北大核心
2018年第6期300-303,312,共5页
Computer Simulation
基金
国家自然科学基金项目(61640302)
青岛市建设事业科技发展项目(JK2015-20)
