A prognostics model for detecting the irreversibility margin of non-reversible electromechanical actuators

This paper presents a prognostics model that was developed for detecting how far an inherently non-reversible ACME screw actuator is from the reversibility limit. The model makes specific reference to an actuator comprised of a dc electric motor, a gear reducer and a power screw with an ACME thread converting rotary into linear motion, but it could be adapted to different actuators configurations. The prognostics model is based on an algorithm that only needs the information on motor current and ambient temperature without the need of specific additional sensors. By comparing the differences between the motor electric currents for two different directions of actuator motion, and fusing that information with those of the components physical characteristics while taking into account the temperature effects, it is possible to detect a decrease of the irreversibility margin and generate an alert when such margin becomes dangerously low.