大型天文光学望远镜超低速跟踪控制

为实现当代大型天文光学望远镜机架伺服系统的高精度控制,利用神经网络预测控制来克服系统中存在的非线性干扰,通过采集机架的输入输出信号训练神经网络来逼近非线性的系统动态,另外,为克服系统外部的风振等非线性干扰,引入了非线性阻尼项来提高伺服系统的跟踪精度.仿真结果表明了该方法的正确性且能获得较高的控制精度.

Ultra-lower velocity control of large-scale optical astronomical telescope

In order to realize high precision control of the mount servo system for modern large-scale optical astronomical telescopes, a neural network predictive control approach is utilized to bate the nonlinear disturbances in the mount servo system. By collecting input and output signals, neural network is trained for approximating the nonlinear system dynamic. Moreover, in view of variable external disturbances like wind buffeting, the nonlinear damp is introduced to enhance the servo tracking performance. Simulation results show that proposed approach is correct and obtains higher control precision.