Cross-Coupled Iterative Learning Control For Linear Motor XY Table

Contour accuracy is one of the most important indexes of XY table motion control system. When we conducts periodic contour machining, the permanent magnet linear synchronous motor XY table could produce two periodic contour errors. The first one is related to processing cycle, the second one is related to position. Iterative learning control could reduce periodic contour errors which have different cycles,so the cross-coupled iterative learning controller was designed, which combined iterative learning control with cross coupling control. Using real-time contour error estimation model estimated contour errors, using iterative learning controller reduced periodic contour errors, according to cross-coupled gain relationship distributed the compensation of XY axis, respectively, the real time compensation of periodic contour errors was completed. However, this method is ineffective for the first period contour error to improve. To improve this method, we could use adaptive robust control to inhibit the contour error for the first period.Firstly, according to the characteristics of linear motor XY table and the need of designing the controller, we analyze the contour error of three kinds processing curves and establish the real-time contour error model for free-form curve. Because the linear motor in the process will be interfered by end effect, load disturbances and uncertain external disturbance, we designed the controller of the linear motor XY table's the single-axis. Displacement of the repetitive controller and PDFF speed controller were adopted to increase system speed, enhance system robustness, reduce single-axis tracking error, finally achieved the purpose to indirectly reducing the contour error.Secondly, for the single-axis controller, it could not completely solve the contour error caused by mismatch between the dynamic biaxial. The cross-coupled iterative learning controller was designed based on PDFF and displacement of the repetitive controller. Using real-time contour error estimation model to estimate the contour error, and using iterative learning control, according to the relationship between the cross-coupling gain to allocate X and Y axis compensation, to achieve the periodic real-time contour error compensation. Theoretical analysis and simulation results show that the solution has a good inhibitory effect to linear motor XY contour errors, but has no ability for the first cycle of the contour errors. Finally, for the iterative learning control and single-axis repetitive control, they have no inhibitory ability to the first cycle contour error. I designed adaptive robust control to reduce contour error effect and enhance its robustness. Real-time contour error estimation model was still used to estimate contour error, and then used cross-coupled iterative learning control to compensate contour error. Theoretical analysis and simulation results show that the solution inhibits the contour error for linear motor XY table effectively.