The Application Of Wavelet Neural Network Based On LabVIEW In Coal Mine Ventilator Fault Diagnosis

Gas explosion is the worst enemy in the safety in coal mine and coal mine ventilation system can prevent gas explosion and exclude gas danger effectively, so it is important to maintain safety production in coal mine by efficient supervising coal mine ventilator.This article is to conduct fault diagnosis of coal mine ventilator, developing a software system in view of LabVIEW, in order to realize online supervision of coal mine ventilator. With the gradually research of this topic, coal mine ventilator fault diagnosis system based on WNN(wavelet neural network) as fault diagnosis model is finally completed. The whole process of research is as follows:(1)Firstly, the article builds a whole hardware structure framework of coal mine ventilator fault diagnosis system; secondly, it designs a WNN model processing vibration signal fault diagnosis The establishment of hardware structure framework and arithmetic model laid a solid foundation to complete the coal mine ventilator system.(2)According to factors of mine environment and own vibration of coal mine ventilator etc., it chooses some hardware devices as the type of AD500T-J sensor with high acuity and quick acceleration and the type of NI PCI-6251video capture card which is used for the core data acquisition equipment.(3)In contrast to traditional methods for signal analysis, it confirms the advancement of wavelet analysis, showing that wavelet analysis has the ability of representing local signal characteristics in frequency domain and can deal with burst signal well. By further analysis of coal mine ventilator failure mechanism and the characteristics of fault frequency, it chooses a modified adaptive learning rate adjustment of algorithm to conduct fault diagnosis on vibration samples. In comparison with BP standard algorithm, the modified adaptive learning rate adjustment of algorithm has a higher prediction error precision and a faster network convergence speed, which has a very good diagnosis effect.(4)Based on LabVIEW as software development platform, MATLAB sets up a software system for assisting developing software which can make basic operations of collecting,magnifying,converting,storing and advanced operations of wavelet denoising filter of vibration signal,time domain and frequency analysis,wavelet decomposition and construction, and uses neural network to realize coal mine ventilator fault diagnosis in final.(5)The developing process of main function modules and its main functions of coal mine ventilator fault diagnosis are researched and explained in detail. Each module interface is well-designed, and comfortable colour assortment and reasonable putting controls are paid attention to, in order to make the whole operating system become more concise,comfortable and hommization.The coal mine ventilator fault diagnosis that is studied in this research brings together several advanced techniques in fault diagnosis field——virtual instrument techniques,wavelet analysis and neural network techniques. It has a positive meaning in the aspects of preventing gas,dust and gas exploration in the pit, maintaining coal mine safety production. The research of the system provides certain theory experience and practice basis for the application of virtual instrument techniques in coal mine security check field at the same time.