火箭炮位置伺服系统的神经网络自抗扰控制  

Active-disturbance-rejection Control of Neural Network forRocket Artillery Position Servo Systems

作　　者：唐冲 童仲志[1] 侯远龙[1] Tang Chong;Tong Zhongzhi;Hou Yuanlong(College of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing Jiangsu 210000,China)

机构地区：[1]南京理工大学机械工程学院,江苏南京210000

出　　处：《电气自动化》2020年第4期115-118,共4页Electrical Automation

摘　　要：针对自抗扰控制器中非线性扩张状态观测器以及非线性控制律参数整定过程繁琐的问题,提出了一种采用BP神经网络整定参数的方法,同时对基于BP神经网络的自抗扰控制器进行仿真分析。结果表明,基于BP神经网络的自抗扰控制器使火箭炮位置伺服系统的稳定性显著提升,在承受外部扰动的环境中仍具有良好的快速性和稳定性,相较于经典自抗扰控制方法有着明显的优势,能够满足工程实际应用中所要求的性能指标。In view of tedious process for setting parameters of the nonlinear extended state observer and nonlinear control law in the active-disturbance-rejection controller(ADRC),a method of parameter setting based on the BP neural network was proposed,and the ADRC based on BP neural network was simulated and analyzed.The analysis results proved that the ADRC based on BP neural network improved the stability of the rocket artillery position servo system remarkably and could maintain good rapidity and stability even under external disturbance,was superior to the classical ADRC obviously,and could meet the performance index required in practical engineering application.

关 键 词：火箭炮 交流伺服系统 BP神经网络 参数整定 自抗扰控制器 

分 类 号：TM57[电气工程—电器]