Afault Diagnosis Model of Marine Diesel Engine Lubrication System Based on Improvedextreme Learning Machine

Abstract

The lubrication system provides lubrication oil to various moving parts in the marine diesel engine. Once faults occurred in lubrication system, it can result in dramatically damage to the diesel engine. Development of fast and accurate fault diagnosis method of lubrication system is therefore highly urged. In this paper, we present a novel intelligent fault diagnosis methodbased on improved extreme learning machine (ELM). Firstly, we use chaotic mapping to enhance capability of the particle swarm optimization (PSO) algorithm; Then, an enhanced PSO algorithm is used to determine initial input weights (connecting input layer nodes and hidden layer nodes) and thresholds of ELM. Finally, we carry out fault diagnosis experiment on the marine diesel engine lubrication system. The experiments demonstrated that the proposed model could achieve more ideal performance.