Research On Iterative Learning Control Of Non-repetitive Disturbance In Linear Servo System

In the high-end manufacturing field,permanent magnet linear motors become the core components of precision platforms due to their fast response speed and high control accuracy,iterative learning control has strong applicability for controlled objects with repetitive running characteristics,which can realize complete tracking of the desired trajectory in a limited time,and is suitable for linear servo control with high speed,high acceleration and high positioning accuracy.However,during the operation of the linear motor,there is non-repetitive disturbance,which will reduce the tracking performance of the iterative learning control.Therefore,suppressing the non-repetitive disturbance is the key to realize the high-performance servo control algorithm.In this paper,the research object is the linear motion platform in the gene sequencing equipment.To solve the non-repetitive disturbance suppression problem,the in-depth theoretical analysis and disturbance suppression algorithm is studied.Firstly,the mathematical model of control object of the system is derived by using linear motor thrust model and platform dynamics model.The model parameters are identified by sweeping frequency experiments,and the control system model is established,which lays a foundation for the research of control algorithms.Secondly,the PD type iterative learning control and filtered type iterative learning control are deeply studied.The convergence,convergence speed and convergence error are theoretically analyzed and simulated.Research focus on the influence of nonrepetitive disturbance on the tracking error of iterative learning control,and the mathematical relationship between tracking error and disturbance is theoretically analyzed and derived.Theory proved to be correct,by comparing calculation results with simulation results.Thirdly,a disturbance suppression algorithm based on wavelet transform and disturbance observer is proposed.The suppression of non-repetitive disturbances from both the iterative domain and the time domain is achieved,and convergence of the algorithm is proved.Simulation results show that the proposed algorithm has stronger suppression ability for non-repetitive disturbances,and it can be applied to iterative learning control to improve the tracking performance.Finally,The linear servo system with PMAC controller,AMC driver and NEWPORT linear motor as the core is built.The effect of the proposed algorithm is verified by experiments.Using S curve as the desired trajectory in linear servo system,only iterative learning control algorithm is adopted,and the maximum error of the uniform velocity section is 0.77 um,while using iterative learning control combined with wavelet transform,the maximum error of uniform velocity section is 0.45 um.When using iterative learning control with disturbance observer,maximum error of uniform velocity section reaches 0.12 um.Furthermore,using iterative learning control based on wavelet transform and disturbance observer,maximum error of uniform velocity section can be reduced to 0.075 um.Expected effect is achieved,which proves the effectiveness and superiority of the algorithm.