Study On The Characteristics Of Electromagnetic Emissions And Its Correlation With Acoustic Emissions During Rock-loading Experiment

As one of the candidate precursory phenomena of earthquake,the anomalous signals of electromagnetic emission before earthquake has always been an important research topic in the field of earthquake monitoring and prediction.The physical mechanism of the anomalous phenomena is still unknown because of its complexity.In order to further investigate the physical mechanism of the electromagnetic anomalies,many scholars from China and other countries have carried out a large of physical simulation experiments in laboratory – rock-loading electromagnetic experiments.The electromagnetic emission signals associated with rock under pressure and deformation are analyzed,which provides a large number of the experimental and theoretical data for exploring physical mechanism of the phenomenon.At present,the exploratory directions of rock-loading electromagnetic experiments mainly focus on the diversity of rock type,the completion of experiment scheme,the improvement of acquisition system and so forth.However,it's limited to the traditional methods in the process of the experimental data,such as the amplitude and frequency of electromagnetic emission signals,the stress and strain during the experiment and so on,and the research on statistical characteristic of the signal is relatively rare.Although some papers analyze the synchronism of electromagnetic emission(EME)and acoustic emission(AE)signals,there's no unified opinion on whether these two kinds of signals are synchronized.Commonly used in earthquake research,The GutenbergRichter(GR)law states the magnitude-frequency distribution of seismic signals.The statistical analysis of b-value based on the GR law is used to study the energy distribution characteristics of EME and AE signals from the failure process of rock.However,some researchers think that earthquake occurrence and rock failure belong to the non-extensive problem,and try to fit the magnitude-frequency relation of seismic,AE and EME signals by the non-extensive parameter q.In recent years,wavelet analysis is widely used in the de-noising of electrocardiogram(ECG),AE,seismic and other signal because of its good time-frequency localization characteristic,but the wavelet de-noising research on EME signals is less,and only in the EME signals from coal-rock.Therefore,this paper mainly carries out relevant research from synchronic relation,non-extensive analysis and wavelet denoising of signal,and the specific research contents and analysis results as follows:(1)The paper checks the synchronism of EME and AE signals produced during rock-loading from the quantity,the arrival time difference and the strength of signal.The results of statistical analysis show that there's a large difference in the quantity of AE and EME signals,the arrival time difference of the corresponding signals is relatively large,and the intensity variation trend of the corresponding signals is not consistent.The two types of signals can't be generated synchronously,and don't belong to the synchronous signals according to the relevant concepts and the above analysis.AE signal is the direct product and EME signal may be the indirect product in the crack growth process,and there're the intermediate transitions between the EME formation and crack growth process.Therefore,it is preliminarily judged that EME and AE signals may come from different radiation sources and have different generating mechanisms.(2)The non-extensive analysis method is used to study the energy distribution characteristics of EME and AE signal from rock fracture and compare with the b-value analysis results,which inferred the type of rock failure and the growth process of crack.In the fitting curves of the energy distribution of EME and AE signals,the fitting results of non-extensive parameter q can better than b-value.The q-values about two kinds of signals were greater than 1,which means that these two kinds of signals originate from the micro-crack nucleation stage and belong to the long-range interaction relation under the non-equilibrium dynamic system.Therefore,the non-extensive parameter q may be more suitable for analyzing the energy distribution of EME and AE signals from rock fracture than the statistical analysis of b-value.(3)Wavelet threshold de-noising method is applied to EME signal generated by rock fracture,which reduces the interference from background noise and highlights the real EME signal.It's necessary to select the optimal wavelet basis function,determine the decomposition layer and design the appropriate threshold function before de-nosing of actual data.The processing results show that wavelet threshold de-noising method can effectively reduce the background noise of EME signal.The threshold function designed in this paper is based on previous research,trying to reduce the deviation between the wavelet coefficients before and after de-noising,and at the same time improving the continuity of the signal.The simulation and actual data processing result show the de-noising effect of the new threshold function is better than other threshold functions.In summary,this paper further broadens the thinking of processing electromagnetic experiment data on the basis of previous research,which makes new attempt and exploration for studying the phenomena of electromagnetic emission from the loading process of rock,so that people have a more comprehensive and deeper understanding of the overall characteristic and the generation mode of electromagnetic emission signals.