基于RBFN的伺服系统逆控制器设计及仿真研究

飞行仿真转台伺服系统中有一些诸如负载变化、机械摩擦等不确定和非线性因素，而神经网络逆控制对模型的不依赖性能较好的解决这些问题。该文给出了转台伺服系统速度跟踪的神经网络逆控制方案，它可以克服转台中负载变化及一些参数变化的影响，且能显著的提高动态精度。通过使用RBF神经网络实现了对象逆动态模型的在线辨识。并直接将该RBFN与PI环节构成一种神经网络逆控制制器，仿真结果表明这种方法具有较好的鲁棒性及较高的跟踪精度，有实际应用价值。

Design and Simulation of RBFN Inverse Controller of Servo System

There are some uncertainties and nonlinear factors such as change of load,mechanic friction in the flight simulator servo system.Because the NN inverse control is independent of the model, it can be used to overcome the uncertainties and nonlinear factors in the servo sysytem.A kind of NN inverse controller for servo system velocity track is put forward in this paper.The controller can overcome the affect caused by the change of servo system load and some parameters.Besides,it can improve the dynamic precision in the low frequency input signal.The RBFN can approach the inverse dynamic model of the object . On-line control of the object is realized by using the RBFN and PI.The analysis and simulation prove that the controller is of good robust performance and high track precision.So it can be applied to the real system.