A new symbiotic evolution-based fuzzy-neural approach to fault diagnosis of marine propulsion systems

This paper presents a symbiotic evolution-based fuzzy-neural diagnostic system (SE-FNDS) for fault diagnosis of propeller-shaft marine propulsion systems. The SE-FNDS combination of fuzzy modeling, back-propagation training and symbiotic evolution function auto-generates its own optimal fuzzy-neural architecture, a significant advantage over previous time-consuming manual parameter determination. Four hundred samples from a test propeller-shaft system are taken over a range of 100-500 rpm, during normal and experimentally induced faulty operation. This database is applied as input/output rule generation and training data for the fuzzy-neural network. Comparison of system construction time and diagnostic accuracy is made by applying the same database to SE-FNDS and four traditional systems. Compared to traditional methods, diagnostic decisions from SE-FNDS show 94.17% agreement with real conditions and less CPU time for system construction. Two nonlinear function approximations are also used to demonstrate the proposed system. The presented design is useful as a core module for more advanced computer-assisted diagnostic systems and for direct application in marine propulsion systems. (C) 2004 Elsevier Ltd. All rights reserved.