Neural Networks Adaptive Sliding Mode Control for Single Gimbal Control Moment Gyro Servo System
Considering the low-speed nonlinearity, uncertainty and large disturbances of the outer gimbal servo system of Single Gimbal Control Moment Gyros (SGCMG), this paper presents a neural networksbased adaptive sliding-mode variable-structure control for higher low-speed system steadiness, robustness and quick response. The main character of this method is that the prior acknowledge of training neural network is not required, also does not need accurate mathematic model of total system including nonlinearity, uncertain parameters and disturbances, Because of the online identification of RBF neural network. The outer gimbal servo system uses a Permanent-Magnet Synchronous Motor (PMSM) driven by sinusoidal wave; the current loop makes use of voltage Space Vector Pulse Width Modulation (SVPWM) method for driving the inverter; and the position loop takes advantage of neural network-based adaptive sliding-mode variable-structure control (NNASMC). The neural networks-based adaptive regulation method is used for improving the system robustness and noise reduction. The simulation demonstrates that a servo system with the NNASMC controller has a higher dynamic and steady precision and better noise reduction, and also eliminates chatter compared with conventional sliding-mode variablestructure controllers.
Zhen Chen Xiang-dong Liu Ya-ping Dai
Dep.of Automatic Control, School of Information Science and Technology, Beijing Institute of Technol Dep.of Automatic Control, School of Information Science and Technology Beijing Institute of Technolo
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）