Research On Friction And Wear Detection And Analysis Technology Of Large Mechanical And Electrical Equipment In Coal Mine

Large mechanical and electrical equipment in coal mines are mostly low speed,heavy load equipment,and the working environment is especially bad.The abnormal friction and wear of parts and components have become a common fault type of this kind of mechanical and electrical equipment.If measures are not taken in time to accurately grasp the friction and wear state of the equipment,once the equipment fails,it will seriously affect the production efficiency,and then bring huge economic losses to the enterprise.In order to solve the above problems effectively,this paper will carry on the in-depth research on the friction and wear state detection and analysis technology of the large-scale mechanical and electrical equipment in coal mine based on the oil detection and analysis technology according to the working characteristics of the large-scale mechanical and electrical equipment in coal mine.The specific research contents are as the following aspects.1)Design and development of oil monitoring system for large mechanical and electrical equipment in coal mine based on Internet technology.The spectral information,viscosity information and iron spectrum image information obtained from oil samples were collected by on-line atomic emission spectrometer,viscometer and iron spectrum analyzer,And the collected oil sample detection information according to the batch number,oil sample number for scientific oil sample management.The system mainly includes: oil sample information management module,iron spectrum image information management module,spectrum information management module,viscosity information management module,equipment detection information online import module,trend automatic analysis module,oil sample detection report analysis module,etc.Based on the above module,the user can give the oil sample detection and analysis report on line,and give the overhaul reminder to the large mechanical and electrical equipment of coal mine using abnormal oil sample.2)The algorithm of friction and wear state analysis of mechanical and electrical equipment based on ferrography analysis is proposed,and the data collected in the designed oil monitoring system is used as the research basis.The algorithm first proposes a multivariate feature extraction algorithm based on the maximum wear particles to extract 22 iron spectrum features from the iron spectrum image.Then,the c-svc classification model is constructed based on the extracted iron spectrum features to classify the two types of iron spectrum images for abnormal wear and no abnormal wear.The simulation results show that the c-svc model without abnormal wear image classification accuracy is 82.5% on average,and the abnormal wear image classification accuracy is 92.5% on average.Compared with the decision tree model,the SVM model based on the maximum wear particle diversity feature and the c-svc model without maximum wear particle diversity feature,the classification accuracy is improved effectively.It provides a theoretical basis for improving the intelligent decision analysis performance of oil-liquid monitoring system of large-scale electromechanical equipment in coal mine,and better realizing the detection of friction and wear state of electromechanical equipment.3)A friction and wear state detection algorithm based on genetic algorithm and iron spectrum analysis is proposed.In order to further improve the classification efficiency of friction and wear state analysis algorithm for electromechanical equipment based on iron spectrum analysis,this paper constructs a diversified feature optimization selection algorithm based on the working principle of genetic algorithm and C-SVC algorithm.The algorithm will optimize the selection of some iron spectrum features from 22 iron spectrum image features,form a new optimization feature matrix,and construct a classification model based on the optimization feature matrix.The simulation results show that compared with the friction and wear state analysis algorithm based on iron spectrum analysis,the total classification accuracy of the model is effectively guaranteed.At the same time,the operation time cost of the model is reduced by about 21.3%,and the calculation efficiency of the original classification model is effectively improved.It provides a theoretical basis for the further optimization of the intelligent decision analysis performance of the oil monitoring system of large mechanical and electrical equipment in coal mines.