煤矿主通风机振动信号特征提取技术的研究

发布时间：2018-04-13 08:49

本文选题：主通风机 + 振动　；参考：《中国矿业大学》2015年硕士论文

【摘要】：煤矿主通风机是保证煤矿安全生产的关键设备,是煤矿井下通风的主要动力来源,通风机的正常工作可以向矿井工作面中输送新鲜的空气,保证井下工作环境良好,所以煤矿主通风机的无故障的连续工作需要得到保证,这对整个煤矿的生产工作都是一种保证。但是煤矿主通风机需要长时间带负荷运转,所以在此长期的运行过程中,由于自身器件和外界因素的影响容易发生一些故障,发生故障的直接表现就是主风机的振动异常,主通风机的振动信号往往包含大量的故障信息,所以随着技术进步和信号处理相关理论的发展,振动信号的特征提取也会有更广阔的发展前景。主通风机的振动信号通常为多分量的非平稳、非线性信号,目前的常见时频分析方法有短时傅里叶变换、Winger-Ville分布、小波变换、EMD等都具有一定的局限性,本文主要采用了局域均值分解(LMD)和小波阈值去噪相结合的方法对煤矿主通风机的振动信号进行特征提取和分析,主要研究内容如下:研究了常用的时频分析方法(短时傅里叶变换、Winger-Ville分布、小波变换)在处理非平稳信号上的表现,分析出各自在非平稳信号分析上的特点和不足。研究了经验模态分解(EMD)和局域均值分解(LMD)的基本原理和基本算法,并在MATLAB平台上进行了仿真分析,并对两种方法在端点处理上进行了比较分析,说明LMD算法更加适应于非平稳信号的分析。研究了小波阈值去噪方法在MATLAB平台上的实现,并与傅里叶去噪进行了对比分析实现,得出小波阈值去噪的优越性,并提出了基于局域均值分解(LMD)与小波阈值去噪相结合的方法对煤矿主通风机的振动信号进行提取分析,并收到很好的效果。
[Abstract]:The main ventilator of coal mine is the key equipment to ensure the safe production of coal mine and the main power source of underground ventilation. The normal operation of the ventilator can transport fresh air to the mine face and ensure a good working environment.Therefore, the failure-free continuous operation of main ventilator in coal mine needs to be guaranteed, which is a guarantee for the whole production of coal mine.But the main ventilator in coal mine needs to run with load for a long time, so in the long running process, some faults are easy to occur due to the influence of its own devices and external factors, and the direct performance of the failure is the abnormal vibration of the main fan.The vibration signal of main ventilator often contains a lot of fault information, so with the development of technology and signal processing theory, the feature extraction of vibration signal will have a broader development prospect.The vibration signals of main ventilator are usually multicomponent non-stationary and nonlinear signals. The current time-frequency analysis methods include short time Fourier transform (STFT) Winger-Ville distribution, wavelet transform (EMD) and so on.In this paper, the local mean decomposition (LMD) method and wavelet threshold denoising method are used to extract and analyze the vibration signals of coal mine main ventilator.The main research contents are as follows: the performance of time-frequency analysis methods (short time Fourier transform Winger-Ville distribution, wavelet transform) in dealing with non-stationary signals is studied, and their characteristics and shortcomings in non-stationary signal analysis are analyzed.The basic principle and algorithm of empirical mode decomposition (EMD) and local mean decomposition (LMD) are studied. The simulation analysis is carried out on MATLAB platform, and the two methods are compared in endpoint processing.It shows that LMD algorithm is more suitable for non stationary signal analysis.The realization of wavelet threshold de-noising method on MATLAB platform is studied, and compared with Fourier de-noising method, the superiority of wavelet threshold de-noising method is obtained.A method based on local mean decomposition (LMD) and wavelet threshold de-noising is proposed to extract and analyze the vibration signal of coal mine main ventilator, and good results are obtained.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD441

【参考文献】