基于改进DRNN的机电作动器PI控制方法研究  

Research on PI Control Method of Electromechanical Actuator Based on Improved DRNN

作　　者：赵世超 赵东标[1] 支程昊 ZHAO Shi-chao;ZHAO Dong-biao;ZHI Cheng-hao(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]南京航空航天大学机电学院,南京210016

出　　处：《微特电机》2021年第12期35-39,共5页Small & Special Electrical Machines

基　　金：国家重点基础研究发展计划项目(973计划,2014CB046501)。

摘　　要：针对机电作动器(EMA)稳定性和响应速度的要求,提出一种基于对角递归神经网络(DRNN)PI控制的EMA速度环控制方法,在运行过程中动态调整增益参数;针对固定学习率无法兼顾系统稳定性和快速性的问题,设计一种自适应学习率算法,增强了控制器的学习能力。仿真结果表明,相较于传统PI控制、模糊PI控制以及固定学习率方法,采用改进DRNN⁃PI控制方法,提高了EMA速度环的动态响应能力和抗干扰能力,能够有效保证EMA整体系统的稳定性和鲁棒性。Aiming at the stability and response speed requirements of electromechanical actuator(EMA),a novel speed loop control method of EMA based on diagonal recursive neural network(DRNN)was proposed.The gain parameters were dynamically adjusted in real time.Aiming at the problem that fixed learning rate cannot strike a balance between the system stability and rapidity,an adaptive learning rate algorithm was designed to enhance the learning ability of the controller.The simulation results showed that,the improved DRNN⁃PI control method improved the dynamic response ability and anti⁃interference ability of EMA speed loop,and could effectively ensure the stability and robustness of the overall EMA system compared with the traditional PI control,fuzzy⁃PI and the fixed learning rate method.

关 键 词：机电作动器 永磁同步电机 对角递归神经网络 自适应学习率 

分 类 号：TM351[电气工程—电机]