Research On Fault Diagnosis Of Drum Shearer

The key equipment in coal mine production is coal mining machine,its structure is complicated,and the underground working environment is bad,so the electrical and hydraulic system failure rate is high,after dozens of coal enterprises in Shaanxi Shenmu City,now that the research of coal mining equipment fault diagnosis is very backward,mainly relying on the traditional experience to judge the problems,and its accuracy efficiency cannot be guaranteed,low degree of automation,has become an important bottleneck in the development of the coal industry,so the study of coal mining machine fault diagnosis method is of practical significance and practical value.This paper mainly studies the overall structure of the shearer fault diagnosis model,neural network,training algorithm and steps,puts forward the optimal learning algorithm(ELM algorithm),the establishment of shearer sub network fault diagnosis model of rolling bearing BP nerve,and study the diagnosis error MATLAB;fault diagnosis research based on hybrid intelligent algorithm of Shearer fuzzy neural network expert the system,establishing the hydraulic shearer samples and model system overheating fault diagnosis of traction,using the single adaptive BP network algorithm and fuzzy BP neural network algorithm for fault diagnosis,and comparative analysis of diagnostic errors and training speed.The main results of this paper are as follows:1.the neural network based shearer fault diagnosis method uses the optimal learning algorithm(ELM algorithm)to avoid the shortcomings of the feedforward neural network algorithm error and weight norm,and improve the generalization of the fault diagnosis network.2.the BP neural network fault diagnosis model of shearer rolling bearing achieves the target error after 6 cycles of training,and the diagnostic error is less than 0.01,which indicates that the BP neural network shearer fault diagnosis method is feasible and efficient.3.the error of thermal fault diagnosis model of the hydraulic traction device of the shearer reaches 0.001.The fuzzy module adaptive BP network algorithm training sample only iterated 1500 times,while the adaptive BP network algorithm has to iterate 3500 times,the former fault diagnosis efficiency is higher.4.the fault diagnosis method of Shearer Based on the hybrid intelligent algorithm is small and the adaptability is better under the same sample training.5.the fault diagnosis method of the shearer based on the hybrid intelligent algorithm optimizes the diagnosis effect of the whole fault diagnosis system.