Research On Fault Diagnosis Method Of Intelligent Coal Sampling System

Coal sample preparation is to analyze or test the coal sample through the steps of reduction,mixing,division and drying,which is used to detect whether the quality of coal sample is up to the standard.Through the integration of coal sample preparation,grabbing and testing,and the introduction of online monitoring and fault diagnosis into the coal sample preparation system,the intelligent development of the coal sample preparation system can be accelerated.Through real-time online monitoring and fault diagnosis,the failure rate of coal sample preparation equipment can be reduced,the performance can be improved,and significant economic benefits can be created for the enterprise.This thesis’ s research object is coal sample preparation system,studies the fault diagnosis method with convolutional neural network or D-S evidence theory,and completes the system monitoring and fault diagnosis design.Firstly,through the analysis of the working principle and technological process of the system,understand the overall architecture,control methods and technical parameters of the system,and give specific explanations on the division,reduction,drying and weighing,for the subsequent monitoring and diagnosis of the coal sample preparation system to pave the way.secondly,according to the working principle,structural module,fault type and fault causes of the coal sample preparation system,the main faults of the system are summarized in three aspects,belt conveyor fault,motor fault and bearing fault.Analyze the causes,manifestations and characteristics of the three types of faults,select appropriate fault detection methods and sensors,and design signal processing circuits.The signal collected by the sensor is transmitted to the upper computer for subsequent fault diagnosis.Thirdly,the thesis makes a detailed study of research methods such as convolutional neural networks and D-S evidence theory,including basic concepts,basic principles and algorithms.Using convolutional neural network method to diagnose bearing faults.The D-S evidence theory is used to diagnose the fault of belt conveyor.The information fusion method of D-S evidence theory is used to extract the feature vector through the wavelet packet analysis,and then the fault diagnosis results are obtained through the information fusion.Data verification shows that information fusion is better than convolutional neural networks to deal with the uncertainty in faults and improves the accuracy of fault diagnosisFinally,The thesis builds a fault diagnosis software platform for coal sample preparation system based on Lab VIEW,uses Lab VIEW visual programming language to collect,record and process the signal of the system,realizes the working status monitoring of the coal sample preparation system,and provides corresponding fault diagnosis result.