基于改进鲨鱼优化算法的倒立摆自抗扰控制研究

Research On ADRC Of Inverted Pendulum Based On Improved Shark Smell Optimization Algorithm

旋转倒立摆是一类单输入多输出的非线性不稳定系统,具有很强的耦合性,传统的控制策略难以满足其控制精度的要求。为实现对倒立摆起摆后在稳定阶段的精确控制,本文采用自抗扰控制策略对倒立摆的摆角和转角进行闭环控制。由于传统的非线性自抗扰控制器具有多个耦合参数,传统的经验整定法难以有效整定,本文通过均匀变异和对立学习策略对传统鲨鱼算法进行改进并对自抗扰控制器参数进行在线整定。通过实验,验证了改进后的鲨鱼优化算法具有较高的全局收敛性,有效的整定了自抗扰控制器参数,提高了控制器精度,使倒立摆获得了良好的稳定性和鲁棒性。

Rotating inverted pendulum is a kind of single-input multi-output non-linear unstable system, which has strong coupling. Traditional control strategies can not meet the requirements of its control accuracy. In order to control the inverted pendulum accurately in the stable stage after swing up, this paper adopted ADRC control strategy to control the swing angle and rotation angle of the inverted pendulum in closed loop. Because the traditional nonlinear ADRC has many coupling parameters, the traditional empirical tuning method is difficult to tune effectively. In this paper, the traditional shark algorithm was improved by uniform variation and random inertia weight, and the parameters of ADRC were tuned online. Experiments show that the improved shark optimization algorithm has high global convergence, effectively tunes the parameters of ADRC, improves the accuracy of the controller, and makes the inverted pendulum obtain good stability and robustness.