Classification of clustered microseismic events in a coal mine using machine learning

Discrimination of seismicity distributed in different areas is essential for reliable seismic risk assessment in mines.Although machine learning has been widely applied in seismic data processing,feasibility and reliability of applying this technique to classify spatially clustered seismic events in underground mines are yet to be investigated.In this research,two groups of seismic events with a minimum local magnitude(ML) of-3 were observed in an underground coal mine.They were respectively located around a dyke and the longwall face.Additionally,two types of undesired signals were also recorded.Four machine learning methods,i.e.random forest(RF),support vector machine(SVM),deep convolutional neural network(DCNN),and residual neural network(ResNN),were used for classifying these signals.The results obtained based on a primary dataset showed that these seismic events could be classified with at least 91% accuracy.The DCNN using seismogram images as the inputs reached the best performance with more than 94% accuracy.As mining is a dynamic progress which could change the characteristics of seismic signals,the temporal variance in the prediction performance of DCNN was also investigated to assess the reliability of this classifier during mining.A cascaded workflow consisting of database update,model training,signal prediction,and results review was established.By progressively calibrating the DCNN model,it achieved up to 99% prediction accuracy.The results demonstrated that machine learning is a reliable tool for the automatic discrimination of spatially clustered seismicity in underground mining.

Seismic modelling of tracked-vehicle signals for monitoring and verifi cation

To better understand characteristics of seismic signals of tracked vehicles measured when passing a sensor line,we numerically modelled force-pulse responses of a layered soil that is similar in its seismic properties to that found at the original measurement site.The vertical-force pulses from the road wheels rolling over the track elements are fitted to the measured ones.Single-pulse seismic waves vary with distance due to diff erent wave types,refl ections at layer boundaries,vehicle velocity and relative position of the left and right track elements.They are computed by a modelling program and superposed at sensor positions with the appropriate slant distance and time shift for each track element.These sum signals are in qualitative agreement with those from the original measurements.However,they are several magnitudes weaker and much smoother.Furthermore,higher frequencies are damped much less at larger distances.Due to the large variability of the sum signals,recognition of tracked-vehicle types exclusively through their seismic signals seems diffi cult.

On the Disadvantages of the NEHRP Soil Classification

The study of the relationship of local ground conditions with the parameters of seismic vibrations carried out by the section of engineering seismology called seismic microzonation. In this branch of applied science radical changes have taken place at the end of the last century. The Commission on Seismic Safety of the National Institute of Building Sciences of the United States has developed new recommendations, which are significantly different from all that used in the world practice of anti-seismic construction. The main provisions of this NEHRP (National Earthquake Hazards Reduction Program) classification adopted in many national building codes, including Eurocode 8. At the same time, a number of papers appeared in subsequent years criticizing the use of the NEHRP soil classification. This article examines in detail and, most importantly, comprehensively the shortcomings of the NEHRP classification.

Distribution Prediction of Middle-deep Lacustrine Source Rocks in Eocene Wenchang Formation in the Kaiping Sag,Pearl Mouth Basin

Oil and gas shows are rich in drilling wells in Kaiping sag,however,large oilfield was still not found in this area.For a long time,it is thought that source rocks were developed in the middle-deep lacustrine facies in the Eocene Wenchang Formation,while there is no source rocks that in middle-deep lacustrine facies have been found in well.Thickness of Wenchang Formation is big and reservoirs with good properties could be found in this formation.Distribution and scale of source rock are significant for further direction of petroleum exploration.Distribution characterization of middle-deep lacustrine facies is the base for source rock research.Based on the sedimentary background,fault activity rate,seismic response features,and seismic attributes were analyzed.No limited classification method and multi-attributes neural network deep learning method were used for predicting of source rock distribution in Wenchang Formation.It is found that during the deposition of lower Wenchang Formation,activity rate of main fault controlling the sub sag sedimentation was bigger than 100 m/Ma,which formed development background for middle-deep lacustrine facies.Compared with the seismic response of middle-deep lacustrine source rocks developed in Zhu I depression,those in Kaiping sag are characterized in low frequency and good continuity.Through RGB frequency decomposition,areas with low frequency are main distribution parts for middle-deep lacustrine facies.Dominant frequency,instantaneous frequency,and coherency attributes of seismic could be used in no limited classification method for further identification of middle-deep lacustrine facies.Based on the limitation of geology knowledge,multi-attributes of seismic were analyzed through neural network deep learning method.Distribution of middle-deep lacustrine facies in the fourth member of Wenchang Formation is oriented from west to east and is the largest.Square of the middle-deep lacustrine facies in that member is 154 km^(2)and the volume is 50 km^(3).Achievements could be bases for hydrocarbon accumulation study and for exploration target optimization in Kaiping sag.