Research On Equipment Fault Classification Method Based On Integrated Incremental Dynamic Weight Combination

With the development of the Internet of Things and information technology in industry,large-scale machinery and equipment in manufacturing industry are constantly emerging in the process of production with a large amount of operation data.By analyzing and extracting equipment fault information quickly and efficiently,and by means of neural networks and other algorithms to achieve effective diagnosis and prediction of fault types,effectively reducing the loss of shutdown or personnel injury caused by equipment failure.Therefore,it has gradually become a research hotspot in the field of intelligent manufacturing.However,in the production process,due to the characteristics of industrial equipment itself,the data generated by it often have the characteristics of unbalanced,high noise and so on.With the increasing of equipment operation data,the accuracy of equipment fault diagnosis model will become lower and lower.In view of the above problems,according to the characteristics of industrial equipment status data and the complexity of industrial production environment,this study proposes a fault diagnosis classification model based on integrated incremental dynamic weight combination,which is as follows.Firstly,by studying the effect of different classifiers on fault diagnosis and classification of bearing equipment in different environments,a dynamic weight combination classification model combining long-term and short-term memory artificial neural network(LSTM)and multi-classification support vector machine(SVM)is proposed to solve the problem of fault feature extraction and classification in high noise equipment status data.Secondly,Based on the LSTM-SVM dynamic weight combination classification model proposed above,the selective ensemble incremental learning mechanism and unbalanced data processing technology are introduced to solve the problem of feature extraction and classification of newly added data and unbalanced sample classes,which are ubiquitous in mass unbalanced equipment status data.Thirdly,the first and second research contents are fused to form a classification model of equipment fault diagnosis based on the combination of integrated incremental dynamic weights(DWCMI).Experiments show that the model can quickly adapt to environmental changes,and can effectively solve the problems of excessive amount of fault data,unbalanced,high noise and unrelated data samples in the process of equipment fault diagnosis.