System and method for detecting and diagnosing pump cavitation

Abstract

The present invention provides cavitation detection systems and methods employing a classifier for detecting, diagnosing and/or classifying cavitation in a pumping system. The classifier can be integral to tie cavitation detection system and/or operatively coupled to the cavitation system via a controller, diagnostic device and/or computer. Parameters such as flow, pressure and motor speed arc measured and/or estimated, and then provided to a classifier system Such systems include Bayesian, Fuzzy Set, nonlinear regression, neural networks and other training systems, for example The classifier system provides a signal indicative of the existence and extent of cavitation. An exemplary classification system is presented that delineates cavitation extent into one or more of the following categories: 0 (no cavitation), 1 (incipient cavitation), 2 (medium cavitation), 3 (fill cavitation) and 4 (surging cavitation). The cavitation signal can be utilized for monitoring and/or controlling a pumping system to mitigate pump wear, failure and other conditions associated with cavitation.

Dependencies are inferred from the claim text.