An Adaptive Correction Method of Diagnosis Experience in Rotating Machinery Fault Diagnosis Knowledge Base

Abstract

The invention discloses an adaptive correction method for diagnosis experience in a fault diagnosis knowledge base of rotating machinery, which collects fault case samples, uses expert fault diagnosis experience to form fault experience samples by a symptom exhaustive method, and combines the fault case samples and fault experience samples together. Constitute the failure sample set. The error back propagation neural network model is built, and the function mapping relationship between the symptoms and faults in the fault sample set is learned through the error back propagation algorithm. The relationship matrix between the fault and the symptom is obtained by the linear expansion method of the learned neural network, and the relationship matrix is ​​normalized to extract the weight coefficient corresponding to the symptom in the diagnosis experience and the reliability of the diagnosis experience. , to realize the self-adaptive correction of fault diagnosis experience in the knowledge base. Using this method, the diagnosis experience can be extracted from the learned neural network model, and the self-adaptive correction of the expert diagnosis experience can be realized. The larger the number of sample sets, the higher the reliability and reliability of the obtained diagnostic experience.

Description

technical field [0001] The invention belongs to the field of fault diagnosis of mechanical equipment, in particular to a method for self-adaptive correction of diagnostic experience in a fault diagnosis knowledge base of rotating machinery. Background technique [0002] The faults of various rotating machinery such as steam turbines, generators, pumps, fans, compressors, and motors are relatively complex, and fault diagnosis mainly relies on expert experience. However, there are certain uncertainties in the experience of experts, and the accuracy of the diagnosis conclusion is greatly affected by the cognitive level of the experts. Experts' experience needs to be continuously supplemented and corrected in combination with the actual failure cases in the project, becoming more and more abundant and closer to the actual situation. This requires expert experience to have adaptive correction ability. [0003] Neural network has strong self-learning ability and is widely used i...

AI Technical Summary

Problems solved by technology

However, the model established by the neural network is essentially a black box model, and the mapping relationship between the output and the input is difficult to express in an easy-to-understand semantic form.

Therefore, this method is mainly used for the adaptive correction of the functional relationship between the input and the output, and there is a big obstacle when applying this method to the adaptive correction of expert experience.

Images