The application of time-frequency domain component method in fault diagnosis of harbor diesel engine

摘要

Time domain signal analysis and frequency domain signal analysis in the machine vibration analyses are two usual methods for the monitor and diagnosis of machine's faults. However, for we often use them separately, we usually get some unsatisfied results. Based on the analysis of vibration signal and considering the nature that the main component can diagnose machine's fault, the text puts forth an improving method, the time-frequency domain main component method, by comprehending the factors of faults in both domains. Therefore, by getting the components in both domains and the correspondent mathematical and physical conceptions, and by using matrix theory and mathematical statistics, we can get the relevant eigenvalue, the transposed matrix, transition matrix and get one component or several ones as the main component. So we can build the corresponding diagnosis model and the mathematical model. Besides these, this paper gives a program frame of getting the main component and the vibration model and presents a basic step of software-making. The method has been used in the harbour fork truck diesel for the vibration signal analysis. By comparing the calculated results with the practical case, the method is satisfied. The main component method can not only predict the fault, but also give the degree of the fault. It provides a convenient method of fault diagnosis for reciprocating machines.