Research On Permanent Magnet Linear Synchronous Motor Control Strategy Based On Backstepping

Permanent magnet linear synchronous motor drive system without mechanical transmission on CNC is more sensitive to model uncertainty and external disturbances. In order to solve the above problems, this dissertation combines some modern control theories and intelligent control methods, with the supports of projects of National Natural Science Foundation of China and Outstanding Talents of Liaoning Province, based on backstepping control, to research the precision positioning control strategy of linear servo system with lumped uncertainties such as parameter variations, end-effects and load disturbances. The main contents are as following:For position tracking of permanent magnet linear synchronous motor servo system, a virtual controller was proposed by constituting a subsystem of position error, followed by a new virtual controller which is designed by selecting the virtual and position errors to constitute a new sub-system, and an actual controller was obtained by revising algorithm of the virtual controller gradually through recursion. Finally the stability of the proposed control scheme was analyzed through Lyapunov stability theory. Simulation results show that the proposed method is effective to restrain the time-varying load disturbances.In order to improve the robustness of position tracking in permanent magnet linear synchronous motor servo system, a position controller is proposed by constituting a subsystem including position error and sliding-mode surface by means of sliding-mode surface function adopting position and speed errors. The stability and robustness of the original controller are ensured. As for the unknown upper bound of lumped uncertainties, an adaptive observer is proposed to estimate it. The stability of the proposed control scheme is analyzed by Lyapunov stability theory. Finally, the feasibility of algorithm was verified by simulations.For exact position tracking of permanent magnet linear synchronous motor servo system, a position tracking controller is proposed by combining backstepping control with intelligent method. A recurrent fuzzy neural network observer is proposed to estimate the upper bound of the lumped uncertainties with parameter perturbation, friction force and external load perturbation. The inputs of recurrent fuzzy neural network are the position error and its derivative, the output is the estimated value of the lumped uncertainties. The output feedback to input by recursive loop of network, the dynamic information of the lumped uncertainties is obtained in real-time. The simulation results show that accurate position tracking was achieved by approaching the upper bound of lumped uncertainties in real-time.The experiment platform of permanent linear motor servo system is built by taking DSP TMS320F2812 as main controller. Put a mass on linear motor mover platform to change the total mass M. Adaptive backstepping and adaptive backstepping sliding-mode control were verified by experiments with rated parameters and mover variation, which shows the effectiveness of theoretical and simulation resultsThe theoretical and practical bases were founded for compensation strategies of linear servo systems to develop practical high-performance and intelligent servo drives.