Application Of Iterative Learning Control To Networked Control Systems With Data Dropouts

In networked control systems,the exchange of information relies on the network to pass between various parts,which increases the uncertainty of the system,and makes it difficult to establish accurate mathematical models.Iterative learning control is a model-free control,which does not depend on a detailed mathematical model of the system.It only needs deviation signals generated by practical and desired outputs to realize the complete tracking for desired trajectories in the finite time interval by simple iterative computation.In this paper,there are two different iterative learning control algorithm for the networked control systems with data dropouts.The main contents of this paper are as follows:Firstly,a stochastic Bernoulli model is established for the networked control systems with data dropouts in both the controller channel and the actuator channel.The sensors and actuators in the system have zero order hold,which can compensate the data dropouts of the system.Secondly,considering data dropouts of the network actuator channel,an open-closed P-type ILC algorithm is proposed for a class of nonlinear discrete systems.A rigorous theoretical analysis of the convergence of the algorithm is carried out.The algorithm uses the previous system error and the current system error to modify the control input,which increases the adjustable gain,makes full use of the historical error information of the system,and improves the tracking performance of the system.Finally,by combining the hierarchical structure of the network control system,a local controller is added to the controlled object.Considering data dropouts of sensor channel and actuator channel,a new open-closed iterative learning control algorithm is proposed for a class of nonlinear discrete systems.A rigorous theoretical analysis of the convergence of the algorithm is carried out.Among them,the local controller adopts an open-loop iterative learning control algorithm to utilize the historical error information of the system;the remote controller adopts a closed-loop iterative learning control algorithm to utilize the current error information of the system.The two controllers work together on the controlled object.Since the local controller does not need to transmit data information through the network,that is,no data dropouts will occur,which can reduce the impact of data dropouts on the control system to a certain extent,and enhances the robustness of the system.