Research Of Mine Main Ventilator Vibration Signal Feature Extraction Techniques

The main ventilator of a coal mine is the key equipment ensuring production safety in coal mines and the main power source in under-mine ventilation. By transmitting fresh air into mining faces, ventilators can guarantee a good under-mine work condition. Therefore, the constant error-free work of ventilators shall be guaranteed, which is also a guarantee to the production work of a coal mine in a whole. However, during the longtime loadingwork of the main ventilator, it is easy to seemly functions under the influence of both internal conditions and outer conditions, whose main case is the unusual vibration of the main ventilator. A lot of breakdown information is included in the vibration signals of the main ventilator. With the development of technologies and theories in signal processing, the feature extraction of vibration signals will have a broader development outlook.Generally, vibration signals of the main ventilator are multi-component, non-smooth and non-linear. Currently, the most common time-frequency analysis methods, including short-time Fourier transform, Winger-Ville distribution, wavelet transform and Ensemble Empirical Mode Decomposition(EMD), all have their own limitations. The thesis uses Local Mean Decomposition(LMD) and wavelet threshold de-noising to extract and analyze the vibration signals of the main ventilator in coal mines. The main content is as following:Studying the performances of common time-frequency analysis methods, namely short-time Fourier transform, Winger-Ville distribution, wavelet transform, in processing non-smooth signals and analyzing their own characteristics and disadvantages in processing non-smooth signals.Studying the fundamental rationales and algorithms of EMD and LMD, conducting simulated analyses in MATLAB, with a comparative analysis of the two methods’ ending processing, and concluding that LMD is more suitable for analyzing non-smooth signals.Studying the application of wavelet threshold de-noising in MATLAB, concluding the advantages of wavelet threshold de-noising through a comparative analysis with Fourier-noising, and conducting an extraction analysis of vibration signals of main ventilators in coal mines through combining LMD and wavelet threshold de-noising, whose effects are satisfactory.