Research Of Iterative Learning Control Algorithm Based On Performance Index Optimization

With the development of modern industrial control,industrial field control tends to be more intelligent,integrated and systematic,iterative learning control(ILC)as a main branch of intelligent control,for the repetitive control tasks to update the control input efforts,the objective of it is to follow a motion profile over a fixed finite time interval to achieve perfect tracking of the control task.Due to the influences of external factors or system actuator constraints,leading to poor performance of tracking control in actual tracking control process.it's of great value and significance to enhance the efficiency and benefits of industrial productions,If we can improve the convergence speed and tracking accuracy by optimizing the ILC algorithms.So in order to improve the system tracking performance,constructed performance index according to ILC learning law and optimizing the performance index and analyzing the relevant parameters based on the theory of optimization,eventually,achieving the goal of enhancing the convergence speed of system output error and improve the actual tracking accuracy.The major research work of this thesis as follows:(1)For the tracking control problem of linear system,an high-order PID iterative learning control gain optimization algorithm with forgetting factor is designed.First,constructing the performance index according to the gain parameter of normal ILC controller,and optimal parameter by performance index.Furthermore,using the error information in repetitive experiences process,creating high-order PID iterative learning controller with forgetting factor,and constructing multiple object parameter optimization performance index for the gain parameter of high-order PID iterative learning controller,by solving the learning gain parameters of high-order optimum iterative learning controller.Finally,the simulation results of the motor driven single mechanical arm control system show the effectiveness and feasibility of the proposed algorithm.(2)For the tracking control problem of discrete linear system with non-repetitive disturbance in output,an iterative learning control algorithm based on the updating reference trajectory is proposed.In order to achieve rapid and high precision output tracking control result at desired points in reference trajectory,the iterative learning controller is optimized by constructing performance index with norm function.Furthermore,when the system output is affected by non-repetitive disturbance in some trials,a new multi-objective performance index function is constructed by Lagrange multiplier algorithm,and the robust iterative learning controller is optimized to improve the convergence speed and tracking accuracy.Finally,the simulation results of the motor driven single mechanical arm control system show the effectiveness and feasibility of the proposed algorithm.(3)For a class of discrete nonlinear repeated systems with random disturbances in input and output,an open-closed loop P-type robust iterative learning trajectory tracking control algorithm is proposed when the initial states is not strictly identical with the given expected value.Based on the ? norm theory,the robust stability of iterative learning algorithm is strictly proved,and the gain matrix parameters of P-type open-closed control law is optimized through the multiple objective function performance,then the convergence of the trajectory output tracking error can be guaranteed under the optimization algorithm,which can improve the convergence speed and tracking accuracy.At last,the simulation results of the mobile robot in two-dimensional motion show the effectiveness and feasibility of the proposed algorithm.(4)For the tracking problem of discrete nonlinear systems with constraints in input,an iterative learning control optimization algorithm based on penalty function and BFGS is proposed.Firstly,the iterative learning controller is designed in terms of BFGS algorithm,and the problem of input inequality constraints is transformed into equality form by adding auxiliary variables.Then,the performance index function is constructed with equality constraints and penalty function,and the parameter factor of the performance index function is optimized to make the output tracking algorithm has the characteristics of monotonic convergence,which aims to improve the convergence speed and tracking accuracy for error.At last,simulation results of the planar motion four-order discrete nonlinear robot arm system show the effectiveness and feasibility of the proposed algorithm.