On condition monitoring of pump pressure in a hydraulic servo-drive system

A novel model-based pump pressure condition monitoring scheme for nonlinear hydraulic systems is presented. A second-truncated Volterra nonlinear model is constructed to describe the characteristic of inherent nonlinearities of the hydraulic systems. It is shown that the Volterra model estimated parameters, based on different input-output measurements, can recognize incorrect pump pressure using a geometric projection scheme. Simulations as well as laboratory experiments shows that any faulty pump pressure conditions can be detected and partially isolated.