基于网格划分策略的自适应ACO算法优化LQR控制器权值

线性二次最优控制（Linear Quadratic Regulator,LQR）是现代控制理论中应用广泛的状态空间设计法,如何选取适当的加权矩阵Q和R的问题一直没有得到很好解决。提出一种基于网格划分策略的自适应蚁群算法来设计LQR控制器,利用蚁群算法获取加权矩阵Q、R及反馈控制率K,并将控制器用于LQR问题的标准试验平台倒立摆系统。仿真结果表明基于网格划分策略的自适应蚁群算法优化设计的LQR控制器超调小,响应速度快,可以有效的对被控系统实施控制。

Controller Weight Matrices Optimized by Adaptive Ant Colony Algorithms Based on Meshing Strategy

LQR（Linear Quadratic Regulator） is the most widely used algorithm to the state-space design in modern control theory,but how to select the appropriate weighting matrices Q and R has never been satisfactorily resolved.This paper presented an adaptive ant colony algorithm based on mesh strategy to design the LQR controller,using the ant colony algorithm to find the LQR weighting matrices and the feedback control rate K,and this controller is the standard test platform for the problem inverted pendulum system.Simulation results show that meshing strategy of adaptive ant colony algorithm to optimize the design of the controller can receive the performance of small overshoot and fast response,and effectively implement control over the system.