Research On Control And Experiment Of High-precision PMLSM Based On Iterative Learning

This thesis is supported by National Natural Science Foundation of China(51075281), Doctoral Program of Higher Education of Special Fund(20112102110002)and Liaoning Province University Innovation Team(LT2010081). In order to meetcontrol requirement of high precision permanent magnet linear servo system, a methodcombining iterative learning control (ILC) with wavelet filter was adopted to reduceperiodic and aperiodic disturbances in permanent magnet linear synchronous motor(PMLSM). The iterative learning controller used for PMLSM servo system obtained fastconvergence and less iteration times by reconstructing input signals and removingaperiodic disturbance components. According to characteristics of PMLSM, theexperimental method was adopted to determine the parameters of L-filter in ILC. Theexperiment results indicate that the proposed method has better tracking precisioncompared with no wavelet filter and traditional PID. The ILC algorithm assures that theoutput trajectory precisely tracks anticipant trajectory with fewer iteration times.Firstly, vector control simulation system of PMLSM was built with simulink toanalyze the relationship between tracking speed and given speed when operating withlow and high speed, as well as under the condition of positive and negative motion. Therelation curves between currents of d and q axes and tracking speed are gained as aresult.Secondly, according to simulink simulation, physical platform of PMLSM was built,after that, analyzing the relationship between tracking speed and given speed andcurrents of d and q axes. The tracking speed curve was gained after analyzing the effectof traditional vector control for motor high speed range with additional periodic load.Finally, the physical platform which was the combination of ILC algorithm andwavelet filter was built based on traditional vector control. The relationship betweentracking speed and given speed and currents of d and q axes were gained when operatingat low and high speeds, as well as under the condition of positive and negative motion with no load. The effect of the combination of ILC algorithm and wavelet filter for highspeed range with additional periodic load was analyzed and after comparing the trackingspeed curve with traditional vector control, the conclusion is that the proposed algorithmis more suitable to control permanent magnet linear servo system.