Robust Control Of Electromagnetic Levitation System Levitation Height Used In Machine Tool

Electromagnetic levitation system has many good quality characteristics such as no pollution, no mechanical contact, no friction, no lubrication, high precision positioning and high feed speed and acceleration. Therefore electromagnetic levitation technology is adopted in gantry NC machine tool to eliminate the friction between the crossbeam and guide, namely the crossbeam is levitated upon the guide without contact in order to achieve finishing-cut.First, on the basis of reading many foreign and domestic references, it summarizes the structure and the principle of operation of the gantry NC machine tool moving crossbeam to analyze the levitation theory. The mathematic model is built. The output feedback adaptive backstepping control is applied in the electromagnetic levitation nonlinear system based on this. Second, the nonlinear model is linearized by using feedback linearization method and linear voltage control and current control model are obtained respectively. On the basis of the model, sliding mode controller is designed, the simulation result shows it has advantage of strong robustness and rapidity but there is chattering as well. In order to eliminate the chattering of sliding mode surface, two-degree-freedom control with sliding mode and neural network is proposed. Namely the input controller is designed based on sliding mode control method and the output feedback controller is realized by a neural network. The simulation result shows the chattering can be effectively weakened by self-learning and treatment capacity of the neural network. Lastly, focus on the disturbance of parameter perturbation, unmodeled dynamics and system quality changed by cutting, the adaptive sliding- H~∞control is designed. The adaptive control is adopted to estimate uncertain parameter to improve chattering. In order to weaken the disturbance, adaptive sliding mode controller with H~∞method is applied to strengthen the system robustness. The simulation results are illustrated that the controller can achieve the accurate orientation with effectively estimate system parameter, little steady state error and strong robustness.