Research On Remote Fault Diagnosis Method Of Mine Belt Conveyor Idler

Carrier idlers,a core part of the mine belt conveyor,which is known as a bond of the mining production system,play an important role in supporting the transmission belt and reducing the transmission resistance.As the key rotation mechanism of the mine belt conveyor,carrier idlers will appear some mechanical faults after continuous operating under high load operation for a long time.Belt tearing and belt burning caused by carrier idlers failures will influence the mine belt conveyor works smoothly and bring a safety threat to coal mining.Not only do long-distance transport,huge amount of carrier idlers and atrocious conditions will cost lots of manpower and material resources for daily audit and scheduled maintenance,but also they can increase the risk of extended service of these parts which have been seriously degraded.Therefore,we need to implement carrier idler remote fault diagnosis for the mine belt conveyor and judge whether this system has fault in time by analyzing these monitoring data,which is conducive to scheduling maintenance plan reasonably and reducing manpower and material resources cost and ensuring that the mine belt conveyor works reliably and safely.Supported by the National Natural Science Foundation of China “Multi-time-scale Hybrid Signals based Health Management of Rotating Machinery Under Varying Operation Conditions”(No.51475455),this dissertation proposes a vibration signal analysis based approach for carrier idler remote fault diagnosis including quantization and bit error characteristics of vibration signals,multiple data stream transmission strategy and fault identification under various working conditions.This novel method is based on measuring technique,signal processing,feature extraction technology and fault identification technology.The corresponding research can provide theoretical support and technical solutions to ensuring the safety operation of the mine belt conveyor.The main contents include:(1)According to signal characteristics of carrier idlers,quantization error model of carrier idler fault signal is established based on uniform quantization mechanism by using Fourier transform,which involves indicators including quantization error,quantization square-error,the first moment of quantization error and the second moment of quantization error,power spectrum of quantization error and theoretical quantified signal to noise ratio.Bit error model of carrier idler is built based on binary coding and Gilbert Elliot model,which involves the first moment of error code and error caused by bit error.Above two models can reveal vibration characteristics of fault signals from carrier idler under the presence of quantization and bit error,which can provide theoretical support for analysis of mechanical fault vibration signals.(2)To solve data transmission problem from multiple sensors mounted on the mine belt conveyor carrier idler,transmission strategies are researched.Based on differences of the sensitivities to the fault signal from multiple sensors,a priority-based multiple data stream transmission scheme is established in the condition of no packet loss.On the basis of analysis of the influences caused by packet loss on the fault signal spectral,a packet loss influence inspired retransmission mechanism is proposed.The problem that how to transmit mass measured data from multiple sensors is solved by using above mentioned comprehensively,which can reduce cost of remote diagnosis and improve remote diagnosis efficiency.(3)On the basis of analysis carrier idler data distribution differences between domains under various working conditions and inspired by transfer learning,fault diagnosis under various working conditions based on domain adaptation using transferable feature and fault diagnosis under variable working conditions based on domain adaptation using feature transfer learning are proposed.Distribution divergences of carrier idler data between domains under various working conditions are reduced,which enhances the generalization ability of fault diagnosis model and improves diagnostic performs under different working conditions.(4)Mine belt conveyor carrier idler experimental platform is set up,which includes experimental platform of carrier idler fault simulator and experimental platform of mine belt conveyor fault simulator.The proposed remote fault diagnosis approach includes three parts previously described is verified and evaluated based on truthful data of carrier idler from above two experimental platforms.The extensive experiment results show that the proposed remote fault diagnosis approach can achieve the mine belt conveyor carrier idler remote fault diagnosis.At last,the work of this dissertation is summarized and the future research direction of the project is proposed as well.There are 130 figures,41 tables and 217 references in this dissertation.