Research On Fault Diagnosis Of Shearer Hydraulic System Based On Information Fusion Technology

In response to the frequent occurrence of coal mine safety accidents and the increasing trend of deep-well coal mining,Xishan Coal and Electricity Group has proposed a fully-mechanized unmanned face construction project.This subject originated from this project.The hydraulic system is an important system for shearer walking,height adjustment,cutting and other operations.Most of the shearer failures are related to it,and it is the key to the intelligent unmanned transformation of the shearer.Based on the information fusion technology,this paper conducts research on the fault diagnosis of the shearer hydraulic system.The main research contents are as follows:(1)The test analysis technology is applied to the construction of the fault diagnosis system of the shearer hydraulic system.Based on the analysis of the typical failure mode of the hydraulic system and its monitoring method,the preliminary plan for the sensor layout of each component of the shearer hydraulic system is determined.,Explore and verify the preset concept,give the expected realization goal,basic principle framework,hardware parameter requirements,and software function logic of the testability design of the fault diagnosis system.The advantages and disadvantages of current fault diagnosis technology are analyzed,and the hierarchical fault diagnosis idea of "time-frequency domain analysis-RBF neural network-D-S evidence theory" is established.(2)Combining the characteristics of real-time condition monitoring data commonly used in the shearer hydraulic system to carry out fault diagnosis,the commonly used monitoring parameters of the shearer hydraulic system such as vibration,pressure,temperature,etc.are analyzed,and the processing methods for different parameter information are given.Through the substantive analysis of the fault diagnosis problem of the shearer hydraulic system,based on the non-linear relationship between the real-time status diagnosis data and the fault mode,the K-means clustering method is used to determine the neural network data center,and the neural network based on RBF is established.The characteristic layer information fusion fault diagnosis model,and further adopting the D-S evidence theory,builds a fault diagnosis information fusion model based on the "time-frequency domain analysis-RBF neural network-D-S evidence theory" multi-theory,which solves the faults from different information sources The fusion of the diagnosis results.(3)Based on the condition monitoring data of double gear pump of MG400(450)/ 930(1030)-GWD shearer hydraulic system in Xishan Coal and power group,the fault diagnosis information fusion model presented in this paper is empirically studied.The results show that the two-level information fusion model for the fault diagnosis of the shearer hydraulic system constructed in this paper can avoid the error of the diagnosis result caused by the neural network fault diagnosis alone,and the accuracy of the fault diagnosis is optimized.It can be applied to the comprehensive mining in the coal mining industry.Unmanned working face construction.