Research On Flexible Iterative Learning Control Based On The Trajectory Primitive Matching And Combining Algorithm

Iterative learning control(ILC)is one kind of control methodology effectively dealing with repeated tracking control problem,which main task is to find the control signal of the desired trajectory in order to achieve perfect tracking in a finite interval and the whole control process need to be completed quickly.But there are still many problems hinder the further application of ILC in the practical engineering.One of main obstacles is required the strictly repeated desired trajectory.And an ILC process needs to renew again as long as the desired trajectory is changed.The preview knowledge of the control singals is no longer useful,which largely degrades the learning efficiency and it is lack of flexibility to perform the task of training.In the paper,the research object is a three-axis feed XYZ platform.It is proposed that how to mine and make full use of the previosly tracked trajectory and the corresponding control singals from the information database based on trajectory primitive matching and combining algorithm.And then the initial control singals of a new trajectory can be obtained directly by using the control singals of the similar trajectory in the information database to solve the problem of fast tracking the new target trajectory.This thesis studies and results are as follows:First,the background and significance of the research is briefly introduced.And the research status of the ILC,the linear matrix inequality(LMI)method and the empirical mode decomposion are reviewed.Second,the whole structure of ILC based on trajectory primitive matching and combining algorithm is introduced.First of all,the definition of the similarity index and the superposition between the two NURBS spacial trajectories is introduced.And then an optimal matching and combining algorithm is desgined under a certain similarity index,which is used to find two or more combined primitive sequences with space patterns similar to the desired trajectory.Moreover,it presents the process of extracting the simailarity trajectory,which goes through affine transformation(rotation and/or transformation)and recomposition.Third,a solution to extract the initial control singals of the iterative learning control is designed.According to the parameters of the affine transformation between the basic trajectories and the similar reference trajectories,the control signal of the similar reference trajectories is obtained by processing,analysis,compensating,transforming the preview control signal of the basic trajectory based on principle of linear superstition.And then the intial iterative control signal of the desired trajectory is extracted directly by using the time-scales transformation between the desired trajectory and the similar reference trajectory.Finally,the validity of the extracted initial iterative control signal is verified by MATLAB simulation.Forth,a kind of segment filter iterative learning control method is designed aim to suppressing the large jump of the recomposed initial iterative learning control singal at the recomposition instants due to the discontinuity.And the controller of ILC is sovled by LMI method.And the cutoff frequencies of the segment filter are obtained by analyzing the time-frequence characteristice of the initial iterative learning control singal.Finally,the simulation is carried out to demonstrate the effectiveness of the proposed method.Fifth,the above methods are programmed in CoDeSys,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.