Research On Vibration-Based Intelligent Fault Diagnosis Of Ship Transmission Shaft

The ship transmission shaft system is a very important part of the ship power plant.The smooth and safe operation of the transmission shaft has a key impact on the safe navigation of the ship and the safety of the personnel on board.Therefore,the research on the fault diagnosis of the transmission shaft system has important practical value for the safe navigation of the ship.Vibration of the propeller shaft during operation will reflect the ship’s fault.Through the study on vibration of the shaft system,sufficient and accurate data are collected for analysis,the working state of the shaft system is identified by means of machine learning,and various faults of the ship’s shaft system are diagnosed.The main work includes the following aspects:Firstly,the relevant theories that need to be applied are analyzed and the theoretical basis of vibration fault diagnosis based on machine learning is elaborated.Secondly,the acquisition and pre-processing of ship fault data is implemented.Through the simulation platform of marine transmission shafts simulation experiment,the research on signal collection and storage of marine transmission shafts is carried out,the data collection is completed by using LabVIEW platform and MySQL database,and the collected data is analyzed by digital signal,so as to verify the accuracy of the data.Thirdly,Then,the fault feature vector extraction model is established.The original fault data is extracted after optimized wavelet packet wt transformation.The dimension reduction algorithm of LDA is used to reduce the dimension of feature samples,and the classifier is used to verify the feature extraction process.Then,the comparative study of BP neural network and SVM support vector machine algorithm in fault diagnosis is carried out.In order to optimize and improve the efficiency of ship transmission shafting fault diagnosis,an improved artificial bee colony optimization support vector machine IABC-SVM algorithm is proposed.Finally,the artificial bee swarm optimization support vector machine IABC-SVM algorithm is applied to the fault diagnosis of ship shafts.An improved scheme is put forward to optimize the support vector machine ABC-SVM using traditional artificial bee swarm algorithm,that is,global search factor strategy is introduced into the search formula of traditional artificial bee swarm algorithm,which solves the problem that it is difficult to select penalty factor and kernel function parameter of SVM support vector machine algorithm.The effectiveness of artificial bee swarm optimization support vector machine IABC-SVM is verified by comparison experiments.