In order to realize the rapid and efficient identification of earthquakes, blasting and collapse events, this paper applies the Convolutional Neural Network(CNN)in deep learning technology to design a deep learning training module based on single station waveform recording of single event and a real-time test module based on multiple stations waveform recording of single event.
On the basis of ensuring that the data is comprehensive, objective and original, the three-component waveforms of the first five stations that recorded the P-wave arrival time of each event are input, and the current mainstream convolutional neural network structures are used for learning test. The four main convolutional neural network structures of AlexNet, VGG16, VGG19 and GoogLeNet are used for learning training, and the learning effects of different network structures are compared and analyzed. The results show that in the training process of various convolutional neural network structures, the accuracy rate and the cost function curve of the training set and the test set of each network are basically the same. The accuracy rate increases gradually with the increase of the training times and exceeds 90%, and finally stabilizes around a certain value. The cost function curve decreases rapidly with the increase of the training times, and eventually the stability does not change near a relatively small value. At the same time, over-fitting occurred in all convolutional neural network structures during training, except for AlexNet. In the end, the cost function of each type of structural training set and test set is finally lower than 0.194, and the recognition accuracy of each type of structure for training sets and test sets is over 93%. Among them, the recognition accuracy of AlexNet network structure is the highest, the accuracy of the training set of AlexNet network structure is as high as 100%, the test set is 98.51%, and no overfitting occurred; the accuracy of VGG16 and VGG19 network structure comes second, and the recognition accuracy of GoogLeNet network structure is relatively low, and the trend curves of the accuracy and cost function in training and test set of each network in the training process are basically the same. Subsequently, in order to test the event discrimination efficiency of the CNN in deep learning in the real-time operation of the digital seismic network, we select the trained AlexNet convolutional neural network to perform event type determination test based on the waveform recording of multiple stations of a single event. The final result shows that the types of a total of 89 events are accurately identified in the 110 events with M ≥0.7 recorded by Shandong seismic network, and the accuracy rate is about 80.9%. Among them, the accuracy rate of natural earthquake is about 74.6%, that of explosion is about 90.9%, and that of collapse is 100%. The recognition accuracy of collapse and explosion events is relatively high, and it basically reaches or exceeds the recognition accuracy of manual determination in the daily work of the seismic network. The accuracy of natural earthquake identification is relatively low. Among the 18 misidentified natural earthquakes, up to 13 events were judged as blasting or difficult to identify due to distortion of waveforms recorded by some stations(They are determined to be explosion and earthquake each by the records of two of the five stations). If sloughing off the recognition type error events caused by waveform distortion due to the background noise interference that overwhelms the real event waveform or waveform drift, the recognition accuracy of earthquake will become 91.4%, and the recognition accuracy of all events will increase from 80.9%to 91.7%, which is basically equivalent to the recognition accuracy of manual judgment in the daily work of the seismic network. This indicates that deep learning can quickly and efficiently realize the type identification of earthquake, blasting and collapse events.

Many small earthquakes occurred intensively and continuously and formed an earthquake sequence after the M_L3.8 earthquake happened at Rushan County, Shandong Province on October 1, 2013. Up to March, 2017, more than 13 000 events have been recorded, with 3 429 locatable shocks, of which 31 events with M_L ≥ 3.0. This sequence is rarely seen in East China for its extraordinary long duration and the extremely high frequency of aftershocks. To track the developing tendency of the earthquake sequence accurately, 20 temporary seismometers were arranged to monitor the sequence activities around the epicenter of the sequence since May 6, 2014. Firstly, this paper adopts double difference method to relocate the 1 418 earthquakes of M_L ≥ 1.0 recorded by temporary seismometers in the Rushan earthquake sequence (May 7, 2014 to December 31, 2016), the result shows that the Rushan earthquake sequence mainly extends along NWW-SEE and forms a rectangular activity belt of about 4km long and 3km wide. In addition, the seismogenic fault of Rushan earthquake sequence stretches along NWW-SEE with nearly vertical strike-slip movement and a small amount of thrust component. Then we apply the P-wave initial motion and CAP to invert the focal mechanism of earthquakes with M_L ≥ 1.5 in the study area. The earthquakes can be divided into several categories, including 3 normal fault earthquakes (0.9%), 3 normal-slip earthquakes (0.9%), 229 strike-slip earthquakes (65.8%), 18 thrust fault earthquakes (5.2%), 37 thrust-slip earthquakes (10.6%)and 58 undefined (16.6%). Most earthquakes had a strike-slip mechanism in Rushan (65.8%), which is one of the intrinsic characteristics of the stress field. According to the focal mechanism solutions, we further utilized the LSIB method (Linear stress inversion bootstrap)to invert the stress tensor of Rushan area. The result shows that the azimuth and plunge of three principal stress (σ₁, σ₂, σ₃) axes are 25°, 10°; 286°, 45°; 125°, 43°, respectively. Based on the stress field inversion results, we calculated the focal mechanism solutions consistency parameter (θ)and the angle (θ₁)between σ₁ and P axis. The trend lines of θ and θ₁ were relatively stable with small fluctuation near the average line over time. Furthermore, the earthquake sequence can be divided into three stages based on θ and θ₁ values. The first stage is before September 16, 2014, and the variation of the θ and θ₁ values is relatively smooth with short period. All focal mechanism solutions of the three M_L ≥ 3.0 earthquakes exhibited consistence. The second stage started from September 16, 2014 to July 1, 2015, the fluctuation range of θ and θ₁ values is larger than that of the first stage with a relative longer period. The last stage is after July 1, 2015, values of θ and θ₁ gradually changed to a periodic change, three out of the four M_L ≥ 3.0 earthquakes (strike-slip type)displayed a good consistency. Spatially, earthquakes occurred mainly in green, yellow-red regions, and the focal mechanism parameters consistency θ was dominant near the green region (around the average value), which presents a steady state, and the spatial locations are concordant with the distribution of θ value. Moreover, all of M_L ≥ 3.0 earthquakes are located in the transitional region from the mean value to lower value area or region below the mean value area, which also indicates the centralized stress field of the region.