Optimization algorithms for prognostics of electrohydraulic on-board servomechanisms

This paper studies the response of an electrohydraulic actuator (EHA) subjected to three different progressive failures (demagnetization of the torque motor, increment of the jack static friction and presence of backlash); in particular, it is focused on the identification of failure precursors able to give an early identification of progressive failures affecting the system, in order to provide tools that can be used to predict its remaining useful life. This kind of analysis belongs to a new discipline, called Prognostics and Health Management (PHM), that focuses on predicting the time at which a system or a component will no longer perform its intended function, estimating its Remaining Useful Life (RUL) and, then, providing an effective diagnostic tool that allows them to exploit a component until it is safe, saving money. In order to conceive an effective prognostic algorithm authors studied the failures effects on the system behaviors, identifying some details in the monitored time-history signals that exclusively got evidence of a particular failure, avoiding confounding each other and allowing pointing out the fault level of the system. For this purpose, the authors developed a new EHA Monitor Model able to reproduce the dynamic response of the actual system in terms of position, speed and equivalent current, even with the presence of incipient faults. Starting from this Monitor Model, the authors propose a new model-based fault detection and identification (FDI) method, based on Genetic Algorithms (GAs) optimization approach and parallelized calculations, investigating its ability to timely identify symptoms alerting that a component is degrading.