摘要
为增强电动负载模拟器的自适应能力以抵抗系统的非线性、时变参数及运动扰动的影响,提出利用小脑模型神经网络(CMAC)与PID的并联进行控制与调节的控制方法。利用PID控制保证系统的初始稳定性,在小脑模型神经网络引入速度信号和误差信号构成二维参考输入,使系统具有很好的自适应消扰能力,减小了多余力矩的影响。仿真证明了该方法的可行性和有效性,收到了很好的控制效果。
A new control structure is presented for load simulator to improve its adaptability and robustness to the influence of nonlinear, time-varying parameters,in which the Cerebellar Model Articulation Controller （CMAC） neural network is adopted together with PID controller. The CMAC can deal with dynamic prob- lem, which is used to identify the generalized inverse model, and the nonlinear problem will be solved prop- erly. The neural network PID controller is designed by requirement for steady. The speed and error signals are taken as reference inputs of the CMAC neural network, which improves the system＇s robustness to the in- fluence of extraneous torque. The simulation, which has a good control effect, proved the feasibility and ro- bustness of the method.
出处
《电光与控制》
北大核心
2009年第9期89-91,96,共4页
Electronics Optics & Control
基金
国家自然科学基金(60874037)
教育部博士点基金(20070287050)
关键词
电动负载模拟器
神经网络控制
智能PID
多余力矩
motor-drive load simulator
neural network controller
intelligent PID
extraneous torque
