Fault diagnosis of belt conveyor gear box based on vibration signal

HU Jing; YUAN Longjiang

doi:10.11799/ce202504018

Abstract：As a critical component of belt conveyor systems,the gearbox's operational status directly determines the reliability and performance of the entire conveying system.Gearbox failures in coal mining operations can lead to production interruptions,significant economic losses,and even personnel casualties.We established a vibration signal-based fault diagnosis model to enhance the predictive maintenance capability of conveyor gearboxes.We took the vibration signal of a belt conveyor as the research object,analyzed the typical fault types,applied variational mode decomposition based on kurtosis permutation entropy evaluation to denoise the vibration signal,established a belt conveyor fault diagnosis model based on least squares support vector machine,and used an improved particle swarm optimization algorithm to optimize the penalty coefficient and kernel parameters of the fault diagnosis model.Finally,the fault diagnosis results were obtained,and the accuracy of the fault diagnosis model was over 95%.Compared with traditional fault diagnosis methods such as SVM,RF,BPNN,the accuracy was increased by 6%,which achieved accurate fault diagnosis,improved fault handling efficiency,enhanced the reliability of belt conveyor operation,and ensured the safety production of coal mines.