环形倒立摆系统改进遗传算法的LQR控制研究  

Research on LQR Control of Improved Genetic Algorithm for Ring Inverted Pendulum System

作　　者：周勇 ZHOU Yong(School of Information Engineering,Engineering and Technology College of Yangtze University,Jingzhou 434020,China)

机构地区：[1]长江大学工程技术学院信息工程学院,湖北荆州434020

出　　处：《机械工程与自动化》2020年第3期31-32,35,共3页Mechanical Engineering & Automation

基　　金：2019年度湖北省教育厅科学研究计划指导性项目(B2019365);长江大学工程技术学院基金项目(2018ky08)。

摘　　要：在一级环形倒立摆系统中,针对线性二次型控制器(LQR)加权矩阵Q和R选择策略的不足,提出一种通过改进遗传算法优化LQR加权矩阵Q和R的方法。该方法选取二次型性能指标J作为个体适应值,在传统遗传算法的基础上通过对交叉率和变异率按S曲线和高斯分布曲线自适应调整等策略来提高遗传算法的全局搜索能力和收敛速度,实现对Q和R矩阵的优化。仿真结果表明:用改进遗传算法优化后的一级环形倒立摆系统LQR控制具有更好的自适应性,系统连杆和摆杆的调节时间分别减少了2 s和1.5 s,有效地提高了系统的稳定性。In the first-order ring inverted pendulum system,in order to overcome the shortcomings of the linear quadratic controller(LQR)weighting matrix Q and R selection strategies,a new method for optimizing LQR weighting matrices Q and R by improving genetic algorithm was proposed.In the algorithm the quadratic performance indicator J was used as the individual adaptive value.Based on the traditional genetic algorithm,the global search ability and convergence speed of the genetic algorithm are improved by adaptively adjusting the crossover rate and mutation rate according to the S curve and the Gaussian distribution curve,and the Q and R matrices are optimized.The simulation results show that the first-order ring inverted pendulum system LQR control optimized by the improved genetic algorithm has better adaptability.The adjustment time of the system connecting rod and swing rod is reduced by 2 s and 1.5 s respectively,which effectively improves the stability of the system.

关 键 词：环形倒立摆 LQR控制器 遗传算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] TP13[自动化与计算机技术—计算机科学与技术]