Mine Main Fan Online Monitoring And Fault Diagnosis

Mine main fan is mainly to provide fresh air for underground workers,to reduce the concentration of harmful gas produced in the underground,and ensure a good working underground environment.If the main fan fails or even causes downtime,it will pose a great threat to coal mine safety.Therefore,the research on on-line monitoring and fault diagnosis of the mine main fan has practical application value.The coal mine main fan is taken as the research object,according to the structural characteristics of the fan,the vibration mechanism of several common faults is analyzed.The monitoring parameters of main fan include bearing vibration and temperature signals,vibration signals are the main parameters for fault diagnosis,and temperature signal plays an auxiliary role in monitoring.The collected data is transmitted through the combination of Zigbee and Ethernet to complete the data monitoring function.Empirical mode decomposition(EMD)is used to decompose the vibration signal.Due to the existence of modal aliasing problem when EMD decomposes the signal with impulse or noise interference,the vibration signal is analyzed by using the ensemble empirical mode decomposition(EE'MD)with Gaussian white noise,and the eigenvalue of the fault is extracted according to the distribution of the intrinsic modal energy.The extreme learning machine(ELM)algorithm is used to fault classify and diagnose of the fan.Aiming at the defect that ELM randomly chooses hidden layer input weights and thresholds,particle swarm optimization(PSO)is introduced to optimize the hidden layer parameters.In order to overcome the problems that network training takes too much time and is easy to fall into a local optimum caused by introducing PSO,the PSO algorithm is improved(IPSO)by adding a variable speed operator and a trigger factor,so that the particle group can shorten the time in the optimization process.The IPSO-ELM algorithm is applied to the fault diagnosis of the mine main fan,and compared with BP,ELM,PSO-BP and PSO-ELM,it is proved that the algorithm is improved in accuracy,stability and speediness.Finally,through the interactive function of the host computer and MATLAB,the fault diagnosis of the monitoring parameters is performed to verify the feasibility of the fault diagnosis algorithm model of this topic.The fault diagnosis of the main fan achieved effectively through deeply studying monitoring technology and fault diagnosis theory knowledge.Some requirements are achieved which are stable information transmission,rapid response and high recognition rate,coal mine safety production is promoted,and there are some values about project application and theoretical research.