Application Of Iterative Learning Control To Networked Control Systems

Iterative learning control(ILC) is an effective technique for control systems with some repetitive nature, which has been widely applied to traditional po int-to-point control systems and introduced in networked control systems(NCSs) gradually. What’s more, ILC is an algorithm which doesn’t depend on detailed mathematical model of the system and only little priori knowledge is needed to acquire the trajectory tracking task in a simple way. So, the purpose of this paper is to study the iterative learning control algorithm with dada dropouts based on the networked control systems. The main contents of this paper are as follows:Firstly, the mathematical model of the data transmission in networked control systems is provided. The data dropout is described as a stochastic Bernoulli process with a given probability and both the measurement signal and control signal data dropouts are taken into consideration. In order to make compensation for the lost data packets, the zero-order holding method is adopted by the data receivers of the controller and actuator.Secondly, a PD-type open loop iterative learning control algorithm is proposed for a class of nonlinear controlled objects in the networked control systems. Owing to the essence of feedforward-based control ILC can revise the control signal in advance. Proportional and differential items of the error signal are adopted to modify the current control signal, thus p roducing new control signal, which takes full advantage of the historical error signals.Finally, a remote open- loop and local closed- loop iterative learning control method is proposed, where the feed- forward controller and feed-back controller act together on the controlled object. The control signal from the remote controller is generated by using error signal of last iteration, which is feedforward control. However, the control signal from the local controller is generated by using error signal of the current iteration, which is feedback control. Besides that, the control system has less dependence on the network because the data transmission between the local controller and the controlled object is without the network.