Research On The Methods Of Detecting And Processing The Grinding Signal

The ball mill is the core equipment for grinding materials such as coals,steels,and metals.Due to the lack of reliable detection methods of the mineral to ball volume ratio(MBVR),the ball mill has worked under conditions of high power consumption and low productivity for a long time.At present,workers mostly control the process of feeding and discharging by listening to the sounds generated from the ball mill,not only resulting in many errors of detection,but also seriously leading to damages to personal health.Practice has proved that the grinding and impact sounds between steel balls,materials and barrels are closely related to MBVR.Therefore,the method of measuring MBVR based on grinding signal has become a research trend.However,the traditional method of detecting and processing the grinding signal cannot effectively remove the noise and extract the characteristics of the grinding signal,resulting in low precision of detection.Therefore,it is of great significance to research on more effective methods of detecting and processing the grinding signal.This topic focuses on the in-depth analysis of the detection method based on the grinding signal's MBVR,and it applies the blind source separation technology,fundamental frequency extraction technology and empirical mode decomposition(EMD)analysis technology to the detection and processing of the grinding signal.Firstly,this paper analyzes the working principle of ball mill and its noise characteristics in detail.Combining with the characteristics that the noise sources of the ball mill are mutually independent and contain information of MBVR to some different degrees,the blind source separation technology is applied to the detection of the grinding signal.The method can separate the grinding signals from the mixed noise signals collected by multiple audio sensors,effectively enhance the SNR of the grinding signals,and lower the level of difficulty in processing the grinding signals.Secondly,this paper concludes that the grinding signal contains the periodic components and the non-stationary components.In view of this feature,this paper proposes a processing method of grinding signal based on fundamental frequency and EMD.The method firstly calculates the fundamental frequency of grinding signal by analyzing the cumulative mean normalized difference function of the signal,and then finds the fundamental frequency band under the respective conditions of empty grinding,normal grinding andfull grinding,finally establishes the relationship between MBVR and the fundamental frequency band of the grinding signal.Due to the fact that the fundamental frequency of the grinding signal cannot react linearly to MBVR when the mill is in normal grinding,this paper applies the EMD technique to the extraction of grinding signal's non-stationary features.In this method,the effective IMF component of the grinding signal is extracted by using the criterion of correlation coefficient and the criterion of discriminative entropy of the IMF energy,and the range of the effective IMF component's energy is found at the normal grinding's different material levels.Finally,the paper uses the two basic features of the fundamental frequency and the energy of the effective IMF component to establish a detection model of the mill's MBVR.Finally,the paper carries out the experiment and analysis of the method proposed in this paper by testing the data collected from the ball mill in the factory.The simulation results show that compared with the traditional method,the proposed method of grinding signal detecting based on the blind source separation can effectively extract the grinding signal and achieves an improvement of SNR by 55%.The proposed method of grinding signal processing based on fundamental frequency and EMD analysis increases the precision of ratio of ball by 64%.This method proposed in this paper provides a theoretical basis for the detection system of ball mill,and has great significances in ensuring the efficient and stable production of the mill.