Air valve fault diagnosis method based on hybrid intelligent technology

Abstract

The invention discloses a gas valve fault diagnosis method based on hybrid intelligent technology, comprising the following steps: S1, using EMD to extract the IMF component of the vibration signal, and extracting the energy and sample entropy of the IMF component containing fault information to form a feature vector; S2. Construct a sensitive feature subset: S21. Calculate the classification accuracy rate of each feature to a single failure mode by CV-SVM; S22. Select the feature with the highest classification accuracy rate as the most sensitive feature; S3. Determine the optimal feature subset: S31. Arrange the failure modes according to the priority from high to low to obtain a set of feature sequences; S32. Eliminate features with low classification accuracy in the feature sequence; S33. Select the subset with the highest classification accuracy from the classification accuracy of the feature subset set as the optimal feature subset. The invention combines intelligent technologies such as EMD and CV-SVM, can effectively identify three types of faults of air valves, has high diagnostic accuracy, and can make accurate judgments on complex types of faults.

Description

technical field [0001] The invention belongs to the technical field of gas valve detection, in particular to a gas valve fault diagnosis method based on hybrid intelligent technology. Background technique [0002] As a general-purpose machine, reciprocating compressors are widely used in petroleum, food, refrigeration and other fields. The air valve is one of the most important parts of the reciprocating compressor. Once the air valve fails, the compressor will lose its ventilation function and cannot work normally. According to statistics, compressor failures caused by air valves account for about 36% of the total number of failures. Therefore, it is extremely important to diagnose air valve failures. The fault diagnosis of the reciprocating compressor valve is essentially a pattern recognition problem, which needs to extract the characteristic parameters of the valve fault signal, and then classify and identify the fault types. [0003] The fault signal of the compressor...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the shortcomings of low diagnostic accuracy and difficult to confirm the diagnosis type in the existing gas valve fault diagnosis, and provide an algorithm that combines EMD, CV-SVM, feature fusion and feature selection, which can identify the gas valve very effectively The three types of faults, the air valve fault diagnosis method based on hybrid intelligent technology, which makes the diagnosis result more accurate

Images