基于广义反向学习的改进鲨鱼算法自抗扰参数整定  

Auto Disturbance Rejection Parameter Tuning of Improved Shark Smell Algorithm Based on Generalized Reverse Learning

作　　者：石春花[1] 刘环 SHI Chun-hua;LIU Huan(Department of Biomedical Engineering, Changzhi Medical College, Changzhi 046000, China;Department of Computer Teaching, Changzhi Medical College, Changzhi 046000, China)

机构地区：[1]长治医学院生物医学工程系,长治046000 [2]长治医学院计算机教学部,长治046000

出　　处：《科学技术与工程》2020年第22期9081-9089,共9页Science Technology and Engineering

基　　金：山西省高校科技开发项目(20091025);长治医学院博士启动基金(BS15015);山西省高等学校科技创新项目(2019L0672)。

摘　　要：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。The parameters of the nonlinear auto disturbance rejection control(ADRC)controller are difficult to tune and greatly affects the control accuracy.Aiming at the problem,an online tuning method of an improved shark optimization algorithm was proposed.Firstly,as the traditional shark algorithm was easy to prematurely converge into a local optimum,and the global search accuracy of the algorithm was low,the shark population was initialized by generalized backward learning,and a non-linear control factor was added during the shark position update process to balance the algorithm global exploration ability and local development ability.Then,Levy mutation mechanism was added in the iterative process to improve the algorithm's ability to jump out of local optimum.Finally,the improved shark optimization algorithm was used to tune the parameters of the ADRC controller online,and the optimized ADRC controller was used in the engineering example for simulation experiments.The experimental results show that the adjusted ADRC controller greatly improves the control accuracy and the anti-disturbance ability.

关 键 词：自抗扰 参数整定 鲨鱼优化算法 广义反向学习 非线性控制因子 Levy变异 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]