A modeling framework for deteriorating control system and predictive maintenance of actuators

Actuators play a central role in industrial automation systems. They are costly, and therefore studying their dependability needs all attention. Usually, an actuator is inserted in a feedback control system, and its mission is to implement a control action delivered by a controller. In this paper, a monotonic actuator deterioration is considered and it is assumed that a relationship exists between the control action and the physical actuator׳s deterioration. A modeling framework is proposed including a non-decreasing stochastic degradation process driving the inability for an actuator to fully implement its role. The prognosis of the actuator׳s residual useful lifetime is derived and used to update the controller׳s setting. The controller reconfiguration completes the maintenance corrective and preventive actions. This new action is suggested as an alternative for maintenance strategy.