基于广义反向学习的改进鲨鱼算法自抗扰参数整定

针对非线性自抗扰控制器参数难以整定、很大程度影响控制精度的问题,提出一种改进鲨鱼优化算法的在线整定方式。首先,针对传统鲨鱼算法易早熟收敛陷入局部最优,且算法全局搜索精度低的问题,通过广义反向学习对鲨鱼种群进行初始化,并在鲨鱼位置更新过程中加入非线性控制因子,平衡算法的全局探索能力和局部开发能力,最后在迭代过程中加入Levy变异机制,提高算法跳出局部最优的能力。其次,将改进后的鲨鱼优化算法对自抗扰控制器参数在线整定,并将优化后的自抗扰控制器用于工程实例中,进行仿真实验。实验结果表明,整定后的自抗扰控制器很大程度提高了控制精度和抗扰动能力。

Auto Disturbance Rejection Parameter Tuning of improved shark smell algorithm based on generalized reverse learning

To solve the problem that the parameters of nonlinear ADRC are difficult to be adjusted, which greatly affects the control accuracy, an on-line tuning method is proposed to improve the shark optimization algorithm. First of all, aiming at the problem that the traditional shark algorithm is prone to premature convergence and fall into local optimum, and the global search accuracy of the algorithm is low, the shark population is initialized through elite reverse learning, and nonlinear control factors are added in the process of updating shark position to balance the global search ability and local development ability of the algorithm. Finally, levy mutation mechanism is added in the iterative process to improve the algorithm The ability to jump out of the local optimum. Secondly, the parameters of the ADRC are adjusted online by the improved shark optimization algorithm, and the ADRC is used in the engineering example for simulation experiment. The experimental results show that the auto disturbance rejection controller improves the control accuracy and anti-disturbance ability to a great extent.