Identification of Typical Fault States of Marine Diesel Engines Based on Optimized BP Neural Network

Abstract

Ship diesel engine is the core equipment of the ship, and its working condition is directly related to the safety and reliability of ship navigation. Once the ship diesel engine fails, it may cause different degrees of sea damage accidents, bringing economic losses and even endangering the life safety of crew members. The fault diagnosis can monitor the state of diesel engine during the operation of the ship and capture the fault signal to ensure that the fault can be found and eliminated in time. Therefore, the fault diagnosis research of ship diesel engine is an important research direction at present. This paper verifies that BP neural network has disadvantages such as inability to escape from local optimal solution and long convergence time, and the BP neural network optimized by genetic algorithm is based on intelligent fault state recognition. The optimized BP neural network has significantly improved in the fitting performance and classification performance. The research results have certain reference value and provide a basis for the research of intelligent fault diagnosis of marine diesel engines.