基于ADRC和RBF神经网络的MSCSG控制系统设计

为了克服外部扰动突变对磁悬浮转子悬浮稳定度和磁悬浮控制敏感陀螺（MSCSG）输出力矩精度的影响，提出了一种基于自抗扰控制器（ADRC）和径向基函数（RBF）神经网络相结合的MSCSG径向偏转控制方法。阐明了ADRC参数对MSCSG控制效果的影响，通过优化设计ADRC，并将RBF神经网络和ADRC结合运用，实现对控制器参数的实时调试，从而克服外界扰动突变的影响。仿真证明所提方法相较于单ADRC控制，不仅改善了解耦控制精度，而且提高了系统对外部扰动和参数变化的响应速度和鲁棒性，可应用于MSCSG的高精度、快响应、强鲁棒控制。 

Design of MSCSG control system based on ADRC and RBF neural network

In order to overcome the influence of external disturbance mutation on the suspension stability of magnetic suspension rotor and the output torque precision of Magnetic Suspension Control Sensitive Gyro (MSCSG), a MSCSG radial deflection control method based on the combination of Auto Disturbance Rejection Controller (ADRC) and Radial Basis Function (RBF) neural network is proposed. The influence of ADRC parameters on the control effect of MSCSG is clarified. By optimizing the design of ADRC and combining RBF neural network with ADRC, the real-time debugging of controller parameters can be realized so as to overcome the impact of external disturbance mutation. It is proved by simulation that compared with single ADRC control, this method not only improves the accuracy of decoupling control, but also improves the response speed and robustness of the system to external disturbances and parameter changes. It can be applied to the MSCSG with high precision, fast response and strong robustness control.