Research And Application Of Reverse Iterative Learning Control Algorithm

In actual industrial processing,there are a large number of systems that operate repeatedly in a limited time interval.And iterative learning control is an effective control strategy for such systems.So far,the researches on the trajectory tracking problem of iterative learning control are to learn directly from the starting point of the desired trajectory to the end point.After this iteration is completed,the system returns to the starting position to start the next iteration.However,there is no relevant research on making use of the features of the desired trajectory.In this paper,from the perspective of effectively utilizing the features of the expected trajectory,a reverse iterative learning control method is proposed for the desired trajectory with mirror symmetry features,the main research work is as follows:Firstly,a mathematical description is given for the trajectory with mirror symmetry properties.By utilizing the inherent features of the desired trajectory with mirror symmetry property,it is decomposed into two independent trajectories by the center point,and the mirror symmetry feature of the two trajectories is used to alternately optimize and adjust the control input of the next iteration period.Therefore,it is not necessary to learn from the starting point of the trajectory every iteration.The convergence speed of the iterative learning control is improved.Then the convergence of the proposed method is analyzed.The reverse iterative learning controller and the traditional iterative learning controller are designed for the linear system.The feasibility and effectiveness of the proposed method are verified by the simulation test of the expected trajectory tracking.Secondly,for a class of nonlinear systems with sinusoidal functions,the simulation tests of the reverse iterative learning control algorithm is carried out.Compared with the tracking effect of the traditional iterative learning algorithm,the feasibility and effectiveness of the proposed method in the symmetric desired trajectory tracking problem of nonlinear systems are further verified.Finally,the reverse iterative learning control method proposed in this paper is applied to the trajectory tracking problem of industrial robots.First,a reverse iterative learning controller based on fixed reference inverse model is designed.Then,for the industrial robot systems with a deviation of the robotic manipulator kinematics model,a reverse iterative learning controller based on adaptive kinematics model is designed.The Kalman filter method is selected to estimate the kinematic model parameters of the robotic manipulator online.And the reference model of the robotic manipulator is continuously updated to reduce the model deviation.The simulation tests and comparison with the traditional algorithm prove that the method proposed in this paper can track desired trajectory quickly.This method has ideal feasibility.It is of great significance for the actual trajectory tracking problem of industrial robots.