The Method Research And Application Of Iterative Learning Control For XYZ Platform

With the improvement of machining accuracy,high request for the control accuracy of XYZ Platform is proposed.For this precision control tasks,iterative learning control is proposed.With repetition work,this control method can realize the complete tracking which feedback control can not reach.Through three axis feed XYZ Platform as research object,they are researched that how to quickly track the new target trajectory and how to realize the complete tracking of the desired trajectory in the presence of the initial position error.This thesis studies and results are as follows:Firstly,through consulting a large number of literature,the methods of iterative learning control and its initial position error correcting are introduced and analyzed.Secondly,the overall framework of the motor XYZ Platform is presented which based on Ether CAT,and the mathematical two-level model of the three axis motor servo control system is studied.Thirdly,a solution to extract the initial control signal of the iterative learning control is put forward,when the current desired trajectory is similar to a trajectory which is formed by matching and combination from former trajectory primitives.The extracted control signals of former trajectory primitives are transformed,such as rotation,translation.The linear interpolation method is used for joining the corresponding control signals before and after trajectory primitives to be satisfied without disturbance during the transition zone.Finally,the initial control input signal of iterative learning control for current desired trajectory can be established.Fourthly,for the problem of position error at each iteration,it is proposed that the segmental filter method of iterative learning control with the initial posit error for XYZ Platform.In the initial segment,the high cut-off frequency is designed to attenuate the initial position error.Then in middle section,the oblique wave transition is set to reduce the generation of disturbance on system.Lastly,in remaining period,lower cut-off frequency to ensure the convergence of learning.Fifthly,the above methods are programmed in Co De Sys,and the experimental results show that the algorithm is effective in practical applications.Finally,the work of this thesis is summarized,and the research works also be prospected in the future.