电动伺服机构扰动补偿与神经网络模糊控制  

Disturbance compensation and neural network fuzzy control of electric servo mechanism

作　　者：胡涛[1] 申立群[1] 曹杰铭 董伟锋 宁佳意 HU Tao;SHEN Liqun;CAO Jieming;DONG Weifeng;NING Jiayi(School of Instrument Science and Engineering,Harbin Institute of Technology,Harbin 150001,China)

机构地区：[1]哈尔滨工业大学仪器科学与工程学院,黑龙江哈尔滨150001

出　　处：《电机与控制学报》2023年第11期10-20,共11页Electric Machines and Control

摘　　要：针对电动伺服机构的模糊控制与扰动补偿问题进行研究,首先依据动力学理论建立电动伺服机构系统模型并利用Simulink软件搭建仿真模型。然后充分分析了系统所受到的摩擦力矩、齿槽力矩、时滞等非线性扰动,设计前馈控制器进行补偿。其次为了进一步改善系统的控制性能,在位置环PID控制器基础上引入模糊控制来动态调整PID控制参数。最后利用BP神经网络实现对量化因子和比例因子的实时整定,改善由于模糊规则及模糊输出论域的不对称性导致在正负行程上效果不一致的问题。从动态响应能力、跟随性能、抗干扰能力、频域响应等方面分别对传统PID控制器、模糊PID控制器和模糊BP网络PID控制器的控制性能进行仿真对比分析,结果表明模糊BP神经网络PID控制器提高了系统响应速度,改善了系统控制品质,可以为航天电动伺服机构结构和控制器设计提供借鉴。The fuzzy control and disturbance compensation of electric servomechanism were studied.First-ly,the system model of electric servomechanism was established according to the dynamics theory and the simulation model was built by Simulink software.Then the nonlinear disturbance such as friction mo-ment,slot moment and time delay were fully analyzed,and a feedforward controller was designed to com-pensate.Secondly,in order to further improve the control performance of the system,fuzzy control was introduced to dynamically adjust the PID control parameters based on the position loop PID controller.Fi-nally,BP neural network was used to adjust quantization factor and scale factor in real time to improve the problem of inconsistent effect on positive and negative strokes due to the asymmetry of fuzzy rules and fuzzy output domain.The control performance of traditional PID controller,fuzzy PID controller and fuzzy BP network PID controller were simulated and compared from the aspects of dynamic response ability,fol-lowing performance,anti-interference ability,frequency domain response,etc.The results show that fuzzy BP neural network PID controller can improve the system response speed and improve the system control quality.It can provide reference for the design of aerospace electric servo mechanism structure and controller.

关 键 词：电动伺服机构 推力矢量控制 扰动补偿 仿真建模 模糊控制 BP神经网络 

分 类 号：TM351[电气工程—电机] TP273[自动化与计算机技术—检测技术与自动化装置]