Researchs On Method Of Early Fault Diagnosis Of Steel Cord Conveyor On The Basis Of EEMD And LSSVM

With the development of technology, steel cord conveyor has the characterof long-distance transmission, large volume and heavy load. Steel cord conveyorhas been the mainly transportation equipment in mining industry, steel industryand metallurgical industry, etc. These industries depend on the growing use ofsteel cord conveyor. Workers are more and more concerned about theoperational status of steel cord conveyor to keep safety.Because of broken wire, deformation, wear and tear, the steel cordconveyor may lead to significant casualties and economic losses when occurringfall or fracture. In order to reduce the occurrence of mine accidents, the articlestates the current development situation of non-destructive testing of steel cordconveyor, then the principle of steel cord conveyor failure and metal magneticmemory testing are analyzed. This article also demonstrates that noise reductionalgorithm to study the metal magnetic memory signal. The advanced andfeasibility least squares supporting vector machine theory is proved with earlyfault diagnosis of the steel cord conveyor.Firstly, the article talks about the non-destructive testing technology’sdevelopment of steel cord conveyor. By analyzing the causes of failure of steelcord conveyor, metal magnetic memory technology principle is discussed andadvantages of metal magnetic memory method in fault detection are stated in thearticle. Compared with the traditional methods, metal magnetic memorytechnology used to detect steel cord conveyor and analyzes the feasibility isbetter. Secondly, the metal magnetic memory signal which is very weak andvulnerable to interference in the field environment will seriously affect the testresults. Based on the distinctive features of ensemble empirical modedecomposition in signal processing, I propose that the improved of empiricalmode decomposition method to reduce noise about metal magnetic memorysignal is much better. This article collected empirical mode decomposition andmetal magnetic memory technology to judge the stress concentration area ofsteel cord conveyor.Thirdly, the multiple features that are extracted from the noise reduction ofmetal magnetic memory signal input to the least squares support vectormachines within early fault diagnosis system. This helps establish the early faultdiagnosis system to recognize and diagnose the operational status of steel cordconveyor.Finally, the particle swarm optimization algorithm to find the optimalparameters of least squares support vector machines. The simulation results,early fault diagnosis system, get the desired accuracy to identify the state of thesteel cord conveyor.