Mechanical wear particle identification method and device, electronic equipment and storage medium

Abstract

The invention discloses a mechanical wear particle identification method, which comprises the following steps of: inputting a training sample of mechanical wear particles into a convolutional neural network (CNN) model for training to obtain a CNN feature extractor for the mechanical wear particles; extracting a feature vector in the training sample according to the CNN feature extractor; inputting the feature vectors into a support vector machine (SVM) classifier for training to obtain a hybrid convolutional neural network model; and through the hybrid convolutional neural network model, inputting the feature vector of the to-be-tested sample of the mechanical wear particles into a trained SVM classifier to identify the abrasive particle type of the to-be-tested sample. In addition, the invention further discloses a mechanical wear particle recognition device, electronic equipment and a storage medium. By the adoption of the method and device, screening of wear particle characteristic information in a manual mode is avoided, and the type of the wear particles can be accurately identified.

Description

technical field [0001] The invention relates to the technical field of on-line monitoring of mechanical system wear status, in particular to a method, device, electronic equipment and storage medium for identifying mechanical wear particles. Background technique [0002] Wear and tear is the main cause of machine equipment failure. As a direct product of wear, the concentration and morphological characteristics of abrasive particles are important reference indicators for wear severity and wear mechanism. Therefore, wear particle analysis has become an important means of equipment wear status monitoring, and is widely used in ships, coal mines, aerospace and other fields. A single or a few abrasive grains cannot truly reflect the wear state of the equipment, and a large number of abrasive grain samples need to be classified to provide a more effective basis for judgment. Therefore, scholars at home and abroad have done a lot of research on abrasive grain type identification...

AI Technical Summary

Problems solved by technology

[0003] However, the existing wear particle recognition method involves a variety of feature vectors such as size, shape, color, and surface texture in the process of feature extraction of wear particles, which is a tedious and time-consuming task. There will also be errors in the process of feature extraction, which will affect the classification accuracy

In addition, there is a problem of information redundancy in the feature information of abrasive grains. For features such as size, shape, color, and surface texture, how should these feature information be screened, how to integrate them, what feature information should be considered as the main factor, and different feature selection Both the method and the fusion method will affect the effect of wear particle identification

Images