Research And Application Of Flexible Iterative Learning Control Method

The target of the flexible iterative learning control(ILC)is to achieve tracking of any desired trajectories accurately and rapidly,and the traditional iterative learning control requires that the desired trajectory must be strictly repeated throughout the iteration process.The key to flexible iterative control is to find the initial control signals of any desired trajectory so that the system can obtain a good tracking accuracy with a fewer number of iterations(even the first time),which avoids the problem that traditional iterative learning control needs relearning because of the change of desired trajectory.The method of the extracting the initial iterative control signals from all previous trajectory tracking control information through matching the desired trajectory is studied in this paper.For the interference caused by the difference in the boundary condition of the initial iteration control signals at the splicing,this paper analyzes the impact of the initial control signal on the system and presents a H_?feedback assistant ILC method.Finally,the proposed algorithm is implemented on the 6DOF industrial robot platform,the experiment results show that the proposed method is effective.The main work and results are as follows:First,the background and significance of the research are briefly introduced.And the research status of the ILC,H_?control and mixed sensitivity problems are reviewed.The whole structure of the flexible ILC with trajectory element matching is described.At the same time,the NURBS model is used to describe the desired trajectory and the trajectory in the trajectory database.Second,the ILC structure of the trajectory element matching is introduced.The NURBS model is used to describe the desired trajectory and trajectory in the trajectory database.By using the optimized matching method based on Kabsch algorithm,a combined trajectory similar to the desired trajectory is obtained from the trajectory database.At the same time,the H_?feedback method is introduced to assist the implementation of ILC,the new desired control signal and the desired trajectory generated in the ILC process for the current desired trajectory will be again stored in the trajectory library,and the trajectory database is constantly enriched.Third,the NURBS model is established to establish the trajectory database,and the optimal matching method of coarse first and then refined is proposed.The RMSE is used to measure the similarity accuracy of two trajectories,and the Kabsch algorithm is used to calculate the rotation translation matrix corresponding to the minimum RMSE of the two trajectories.First of all,by reducing the matching precision requirement and increasing the search step,we can getsome whole trajectories which has more similar segments with the desired trajectory from the large number of trajectories.Then,several trajectory elements similar to the desired trajectory segments are obtained by improving the matching precision and shortening the search step.Finally,the some trajectory elements are combined into a combined trajectory similar to the desired trajectory by rotation and translation.Forth,the method of extracting the primary control signal is studied,and the influence of the initial control signals on the initial output error of the system and the influence of the control signal on the output of the system at the jump are deduced mathematically.The structure of H_?feedback-assisted iterative learning control system is designed,and the process of proving the convergence condition of the system is introduced,at the same time,the controller parameters of the entire system are designed by selecting an appropriate weight function.By comparing the initial control signal extracted from this paper with the ILC control method with zero initial control signal,the validity of the initial control signal extracted in this paper is verified.Fifth,basebased on the CoDeSys software development platform to construct the ER50 six axis industrial robot kinematics model,realize the power-on and back-to-origin functions of the robot,Implementing feedback-aided ILC algorithm through programming.The experimental res ults demonstrate the effectiveness of the proposed method.Finally,the work of this thesis is summarized,and explain some deficiencies and put forward some prospects.