Lnear Servo Two Position Loop Cross-Coupling Synchronous Control

Based on the National Natural Science Foundation Project named "Research on Dual Linear Motors Synchronous Micro Feed Technology of Gantry Moving Type Computerized Numerical Control Machine (No. 1071201-1-00)", the problem of the linear servo dual position loops synchronous drive control is researched in the paper. For overcome the impact on the control performance produced by non-linear, strong coupling, uncertainties and other factors. Studied the simultaneous transmission of two-position loop controller based on conventional fuzzy method, self-adjusting quantization factors, scale factors fuzzy control method and fuzzy adaptive PID approach.First, reviewed the development of linear servo system on the basis of reading a large number of domestic and foreign-related literature. Introduced the principle of vector control and permanent magnet linear synchronous motor model, on the basis of analyzing basic structure and working principle of permanent magnet linear synchronous motor Finally, overview the principles and classification of synchronous control and factors that affect the synchronous control.Second, for single-axis servo system, designed two sets of control program. First, using PI speed controller, feedforward control and sliding mode feedback control strategy. Second, designed PDFF speed controller in order to improve single-axis system anti-jamming capability and rapid response capability.At the same time, for the non-synchronization problem, designed cross-coupling fuzzy adaptive PID controller. Fuzzy adaptive PID controller combines the advantages of fuzzy control and PID control to adjust PID parameters on-line. Designed fuzzy neural network self-organized cross-coupling controller, the controller directly achieve fuzzy quantification, fuzzy rules fast parallel reasoning operations, inverse fuzzy and the average sentence. PNN network as performance measurement indicators network control the amount of correction. Through off-line training, made PNN associative memory performance measurement decision table in order to provide FRNC network on-line learning signals, and use Kalman filter learning algorithm which has fast convergence and good numerical stability.In the end theoretical analysis and simulation results show that the designed control strategies have robustness, fast response and the dynamic process synchronization error is small, they are all able to meet the accused objects to the requirement of high-precision synchronous control.