Iterative Learning Control Of Direct-driven XY Table Based On Empirical Mode Decomposition Algorithm

Industrial processing technology is an important index of national productivity and industrial strength. Recently the numerical control machine has always drawn attention because of its capacity of fast and accurate processing. Permanent magnet synchronous motor has small size, high efficiency, and good dynamic performance. Linear motor has simple structure, high precision and, high speed. Permanent magnet linear synchronous motor which combines double advantages has become the prevailing drive motor of the high-speed and high-grade CNC. In addition, the PMLSM driven XY table which used to achieve positioning and feeding of planes is widely used in NC milling machine, boring machine, drilling machine and machining center. The precision of motion control depends on the servo control structure and the mechanical structure, due to the mature mechanical structure of numerical control equipments, it is necessary to design a good controller to improve precision.Firstly, this thesis introduce the structure, working principle and application of the permanent magnet linear synchronous motor and direct-driven XY table, analyze the disturbances of system, and establish the mathematical model of them. And also define the tracking error and contour error, and deduce the contour error expressions of straight, arc and arbitrary orders respectively.Design the cross-coupled controller by ILC algorithm can reduce the contour error on biaxial system gradually through repetitive control, but there is an inherent problem in the ILC. According to the above problems, this paper conducted the following researches.Secondly, traditional control methods have limitations on the uncertain, time-varying and nonlinear systems, to achieve accurate tracking control of the PMLSM servo system, this paper design the iterative learning position controller and PI speed controller. The design of traditional cross-coupled controller and uniaxial feedback controller is separated, lacking systematization, to achieve accurate and coordination contour control of the direct-driven XY table, this paper establish the uniaxial feedback control system and the biaxial cross-coupled ILC structure simultaneously, and present the expressions of the controller parameters directly refers to transient response. ILC uses the error and past control input to calculate the current control input, so the accumulative error will directly affect the convergence speed and the stability of system, even can lead to divergence. To solve this problem, this paper decompose the tracking errors in learning process using empirical mode decomposition algorithm, exclude the divergent components, and improve the convergence speed, tracking accuracy of ILC system and contour accuracy of cross-coupled ILC system.Finally, build the simulation systems of permanent magnet linear synchronous motor and direct-driven XY table under the Matlab/Simulink, program to realize ILC, verify the effectiveness of ILC and cross-coupled ILC controllers. Use Matlab programme to realize the empirical mode decomposition of errors, eliminate the divergent components and run the system again. Analysis and compare these results of simulation, validate the effectiveness of these new methods.