Research Of Four-axis Iterative Learning Control For Homogeneous Trajectory Group

There are many vary-trajectory tracking problems in industrial processes.In these processes,the output trajectory is required to track the various trajectory.For tracking same trajectory,iterative learning control(ILC)can show high performance.However,for tracking changing trajectory,traditional ILC can hardly tackle it,because the previous control data cannot be inherited.Based on this,the study of vary-trajectory tracking problem has very important application value.For the tracking problem of a kind of homogenous trajectory group(HTG),four-axis iterative learning control(4A-ILC)based on knowledge inheritance is proposed.Firstly,the reference control data from the initial trajectory via ILC is obtained.Then,the control date for the next trajectory is got by the essential feature transformation when the system firstly operates.Finally,the corresponding convergence conditions of the proposed algorithm are given.The main contents of the paper are as follows:(1)HTG is defined by the essential feature for the study of vary-trajectory.HTG is derived from the initial trajectory through the essential feature transform(ie,gain transform and offset transform).All trajectories of HTG have the identical essential feature and possess the characteristics of the gradual amplitude.(2)4A-ILC is proposed for the tracking problem of HTG.4A-ILC contains the time-axis,iterative-axis and amplitude-axis that the traditional ILC has,as well as the trajectory-axis it does not exist.4A-ILC processes the control data of the trajectory one by one to realize the knowledge inheritance,so as to reduce the overall tracking times of HTG and improve the learning efficiency.Finally,the validity of 4A-ILC is verified by theoretical analysis and numerical simulation.(3)For a kind of system with double noise(both intra-system-noise and outside-system-noise),the four-axis iterative learning control based on double Kalman filter(DKF-4A-ILC)is proposed.Firstly,the Kalman filter in time-axis is to suppress intra-system-noise to reduce tracking error.Secondly,the Kalman filter in iterative-axis is to suppress outside-system-noise to further reduce tracking error.Then,knowledge inheritance in trajectory-axis realizes the fast tracking of HTG via DKF-4A-ILC.Finally,theoretical analysis and numerical simulation verify the effectiveness of DKF-4A-ILC.