Gear system multi-fault diagnosis method based on COM assemblies

Abstract

The invention discloses a gear system multi-fault diagnosis method based on COM assemblies, which integratedly utilizes empirical mode decomposition, wavelet threshold noise reduction, a higher-order cumulant theory and a COM assembly technology. The wavelet threshold noise reduction directly acts on a high-frequency intrinsic mode function component obtained through the empirical mode decomposition rather than acting on a reconstruction signal obtained through a whole signal. Empirical mode decomposition-higher order cumulant processing are performed on the reconstruction signal after the noise reduction, and according to a spectrum analysis result, a diagnosis about a fault mode and a damage degree is made, such that the diagnosis precision and efficiency can be enhanced, and diagnosis functions are enriched. During an implementation process, the COM assembly technology is taken as a software realization means of a diagnosis system, and each part is developed to a COM assembly and can be combined to form an application system which can carry out fault mode and damage degree diagnosis; and by using the method provided by the invention, typical single-fault diagnosis can be carried out, composite multi-fault diagnosis can also be carried out, and the development and online upgrading of the diagnosis system are facilitated.

Description

technical field [0001] The invention belongs to the technical field of mechanical fault diagnosis, and in particular relates to a multi-fault diagnosis method of a gear system based on a COM component. Background technique [0002] With the increasing perfection, complexity and automation of mechanical equipment, the theory and method of equipment fault diagnosis have been widely researched, developed and applied at home and abroad. [0003] Invention patent CN102122133A discloses an "adaptive wavelet neural network anomaly detection fault diagnosis and classification system and method". The adaptive wavelet neural network of the system can automatically establish an adaptive mechanism for the samples to be detected. The feature information of the signal is extracted, and more accurate abnormality detection, fault diagnosis and positioning results can be obtained. In the patent CN102288286A, "a vibration acceleration sensor gearbox measurement point accuracy analysis and ev...

AI Technical Summary

Problems solved by technology

[0005] At present, the mechanical fault diagnosis device or system developed based on vibration testing has simple theory and single function, and the disadvantage is that it cannot be implemented online; Research results; the diagnostic effect on multiple faults and damage is poor; the diagnostic accuracy is not high, and even misjudgment occurs; it is difficult to upgrade the diagnostic software

Images