A data dropout compensation system based on iterative learning control techniques

This paper deals with iterative learning control (ILC) systems subject to the presence of dropouts. The system is composed by a set of discrete-time multiple input-multiple output (MIMO) linear models, each one with its corresponding actuator and its sensor. The ILC law is processed in a control device sited far away of the iterative models so the control signals have to be transmitted from the controller to the actuators. Moreover, the ILC law uses the measurements of each iterative model to generate the inputs vector to be applied to its subsequent iterative model so the measurements of the iterative models have to be transmitted from the sensors to the controller. An unknown set of such transmissions is susceptible to suffer failures at any time instant. A compensation dropout technique is used to replace the data which have been lost in the transmission processes with the aim of guaranteeing the convergence of the output errors.