磨音信号检测与处理方法研究

Research on the methods of detecting and processing the grinding signal

摘要:

针对传统磨音检测和处理方法无法有效去除噪声和提取磨音特征，导致球磨机料球比检测精度较低的问题，提出了一种基于盲源分离(BSS)的磨音检测方法和基于经验模态分解(EMD)的磨音处理方法。该方法首先利用BSS技术从多个音频传感器采集的混合噪声中分离出磨音信号，然后对其进行EMD分析，利用相关系数准则和本征模函数(IMF)能量的判别熵准则提取了磨音的有效的IMF，最终找到了不同料球比水平下磨音有效IMF能量的范围。实验结果表明提出的磨音检测方法可有效提取出磨音信号，提出的磨音分析方法大大提升料球比检测精度,为应用于实际的球磨机料球检测系统提供了理论基础。

Abstract:

Based on the fact of that the traditional methods of detecting and processing the grinding signal cannot effectively remove noises and extract the characteristics of grinding signal, a method of detecting the grinding signal based on blind source separation(BSS) and a method of processing the grinding signal based on empirical mode decomposition (EMD) is proposed. The method firstly uses BSS to separate the grinding signal from the mixed signals collected by multiple audio sensors, and processes the grinding signal using EMD, then extract the effective intrinsic mode function (IMF) of the grinding signal using the criterion of correlation coefficient and the criterion of discriminative entropy of the IMF energy. Finally, the range of the effective IMF component’s energy is found at different material levels. The experimental results show that the proposed method of detecting the grinding signal can effectively extract the grinding signal, and the proposed method of processing the grinding signal greatly improves the detection precision of the material to ball volume ratio, providing a theoretical basis for the detection system of ball mill.