Low-speed Tracking Control Of The Simulator By The Immune Radial Basis Function Network

Electric-driven simulator tends to be unstabe due to the disturbance moments like friction moment when it is running in low-speed, which impacts the control performance of the simulator seriously, therefore, how to overcome the nonlinear problem mainly caused by the friction is becoming to be an important part in the control area of the flight simulation servo system.In this paper, ulitlize immune algorithm (IA) and radial basis function (RBF) network, in order to investage the properties and compensational control problems of the friction of the simulator with traditional PI controller. Firstly, use the describing function method to investigate the stability problem of the simulator based on the ideal simulator model and the subsection friction model, analyze the dynamic-static properties and the action mechanism of the friction in the simulator, compare two important friction models in the problem of the precision and the effectiveness; Secondly, a control method based on friciton model is proposed: build a nonlinear model based on the LuGre friciton model of the simulator, use the immune algorithmm to identify the parameters of the model, transform the model into the inverse model, based on which the forward control loop is realized; Thirdly, a control method based on non-model is proposed: identify the inverse model the simulator by RBF network, based on which the forward control loop is built; Lastly, design a new RBF network with smaller strcture and higher genelization by IA, identify the inverse simulator model, based on which the forward control loop of the system is built.The simulations and experiments results of the control for a one-axis simulator indicate that the two presented compensation methods can well overcome the nonlinear problem mainly caused by the friction, so that the control performance of the simulator is improved.