Research Of Unattended Engine Room Automation Monitoring And Fault Diagnosis

Marine engine room is the fountainhead of the marine power plant. Its normaloperation has the important significance to the ship transportation safety. The engine roomautomation monitoring and fault diagnosis technologies are primary parts of modern shipwork safety.Focusing on diesel engines, this thesis attempts to do a study on modulardesigns of engine room monitoring, faults feature extraction, fault diagnosis method ofdiesel engines and Fuzzy Neutral Network optimization by applying principles of AntColony Algorithm, Virtual Instrument and test technology.This paper aims to research the remote fault diagnosis to marine diesel engines usingLabView. Moreover, applying modular thinking in Virtual Instrument technology, the thesisdesigns engine room monitoring and fault diagnosis technologies to overcome theshortcomings of traditional development environments (VB, VC++, C, etc), such as longdevelopment cycle, limited running speed and difficulties in debugging, maintaining andredundancy. The data acquisition, processing, central display and alarm-control are alsorealized, forming solid foundations for the diesel engine remote fault diagnosis.In recent years, with the swift advance in breakdown diagnosis technology, instead ofsimple corresponding relationship, complex relationship often exists. Fuzzy NeutralNetwork (FNN), formed by combining fuzzy logic and artificial neutral network, couldeffectively apply fuzzy information. With special features of self study, FNN could betterhandle the task of fault diagnosis. This thesis combines the latest Ant Colony Algorithm andFuzzy Neutral Network to achieve the optimization training of the parameters and theoptimization of network weight and threshold values, leading to better convergenceperformance and speed. Avoiding shortcomings of local extremum and slow convergence innetwork training, at the same time, simulating in the environment of MATLAB, the paperdemonstrates considerable feasibility and reliability.