The Js-type Rotary Kiln Exhaust Fan Vibration Signal Processing And Fault Diagnosis

It is the key equitment in the system of rotary kiln of the Great Wall Aluminum Industry Company to the type JS- discharge ventilator, the condition monitoring and fault diagnosis to it has already become necessary link in alumina production. If the fault has emerged in some place of the ventilator, we have not controlled and eliminated in time ,it may cause not merely the damage of the equitments and the working environment,even cause the comsequence of injury and death of workers. In order to guarantee the normal running of the equitment, entrusted by the Great Wall Aluminum Industry Compan,we have developed the system of the condition monitoring and intelligent fault diagnosis to the discharge ventilator,the system has been realized to the discharge ventilator monitoring and controling in real time, trendency prediction, fault diagnosis and so on, the system is runing well at the sence of the company at present.In this thesis, I firstly have studied that the useful signal's abstraction problem from vibration signal including the bigger noise of the discharge ventilator under the complicated working congdition,mainly researched scalping false points methods, tendency part canceling methods, frequentcy mixture and energy leaking problem, digit filter design and so on, especially explorated the problem of the tendency part canceling ,constitutes a kind of general matrix representation means of the trendency part,and simulated running state of the discharge ventilator in factory, have proved to the procedure that is designed through the emulation experiment.Secondly, vibration signal of the discharge ventilator wavelet elimination method has been studied by applying wavelet analysis method,in the thesis, on the basis of analysing the wavelet characteristic of the vibration noise signal and the useful signal of the ventilator, relatively and profoundly studied noise elimination technology of wavelet forced-valve, hard-valve and soft- alve, especially explorated soft-hard valve compromise means wavelet elimination technology's application in vibration signal treatment process of the discharge ventilator wavelet,and through emulation compared result with different wavelet function, noise elimination means,finded out suitable wavelet function and wavelet noise elimination means ,confirmed relevant parameters.Again in the thesis I have analysed the vibration mechanism of the discharge ventilator's typical fault, summarized the basic vibration characteristic of the faultdiagnosis method,studied the basic principle and algorithm that the neural network technology is used in the intellectual fault diagnosis, especially explorated the parameter confirm method standardized on least step length to confirm BP neural network model,by neural learning and training,constitutes the fault diagnosis model suitability for discharge ventilator's character, and use the on-the-spot fault signal to carry on the fault diagnose experiment.Finally, in this thesis I also have analysed and designed the condition monitoring and fault system diagnosis of the discharge ventilator,andanalysed and introduced the system hardware composition principle, measurement-pot's distribution, the instrument and equipment's chosen, software design procedure, the main interface and applying result.