Based On The Study Of STM32 Mine Belt Intelligent Fault Diagnosis System

As one of the important equipment in the coal mine transportation system,the belt conveyor directly determines whether the coal mine output can meet the demand.However,due to the very harsh working environment in the underground coal mine,it is inevitable to cause wear and tear on the belt conveyor and other equipment,and it is not convenient for manual maintenance due to the huge belt conveyor transportation system and the very long belt transportation distance.So as to make the belt conveyor in the process of high strength,long run will inevitably happen tear,wandering,broken belt,skid and a series of common sudden fault,and also has a strong correlation between failure,once a failure occurs often unable to find the fundamental problem and can not get timely maintenance,resulting in serious consequences and even security accident,brings to the coal mining enterprises and mine staff cannot make up for the loss.Therefore,it is urgent to design a set of intelligent fault diagnosis system with accurate detection,fast response,safety and reliability.Most of the traditional fault diagnosis technologies adopt single sensor diagnosis and improve the accuracy of diagnosis results by improving the accuracy and sensitivity of sensors.However,with the continuous progress of science and technology,the scale of belt conveyor transportation system is expanding,and the coupling degree between fault diagnosis systems is also increasing and the correlation between faults is further improved.Therefore,it is often not reliable to judge the fault cause through the unilateral information collected by a single sensor.Information fusion technology,multi-sensor real-time acquisition at the same time,the different types of information on the space,and through certain rules of fusion,so as to make the diagnosis results fully take into account any may cause the fault information,effectively avoid the diagnostic results due to incompleteness of gathering information is not accurate,so as to improve the reliability of fault diagnosis system.The classical d-s evidence theory,as a reasoning algorithm to deal with uncertain information,can well combine the different levels of information collected by multiple sensors organically through fusion rules,and to some extent improve the diagnostic accuracy.However,due to the influence of objective factors,the information collected by multiple sensors is often in conflict.At this time,due to the limitations of the classical d-s evidence theory in handling conflict evidence,the fusion result is greatly different from the objective reality.Based on this,the main reason for the failure of the classical d-s evidence theory is that the conflict evidence is not fully considered and the fusion result is inconsistent with the actual situation.And through the analysis of the relevant improvement methods,put forward by introducing concept of relevance of evidence and global trust makes the improved D-S evidence theory can fully take into account including conflict evidence within the credibility of any evidence,thus effectively avoid the diagnosis due to the incompleteness of information is not accurate,and three kinds of common faults in the operation of mine belt conveyor probability model is established and numerical calculation,analysis shows:When the collected evidence has conflict,the improved algorithm can process the conflict evidence more quickly and accurately than other improved methods,so that the fusion results are more consistent with the actual situation.Finally,the upper computer monitoring system is designed through MCGS configuration software to detect the running state parameters of each equipment of mine belt conveyor in real time and store relevant alarm information,so that the staff can quickly know the location and type of fault occurrence through the man-machine interface and timely take reasonable measures for maintenance.