Abstract:In the process of automatically identifying infrasonic waves in a focal region prior to the occurrence of an earthquake, tripartite array arithmetic has difficulty locating the source of infrasonic waves as it cannot automatically screen and identify a large amount of abnormal-infrasonic-wave data. This leads to low monitoring accuracy and efficiency prior to an earthquake. In this work, we present an automatic method for identifying anomalous infrasonic waves in the cloud computing environment. We constructed a JNS abnormal-infrasonic-wave data acquisition and screening module to scan the access port in real time, and quickly provide feedback regarding abnormal infrasonic data. We use an NDS abnormal-infrasonic-data-sequence detection algorithm to quickly identify a wrong sequence matrix, and accurately retrieve, locate, and lock the abnormal infrasonic wave data. This automatic recognition method can be used to classify anomalous infrasonic waves and determine whether the seismic signal is suspicious. The experimental results show that the proposed method can efficiently and automatically identify abnormal infrasonic signals prior to an earthquake, with a maximum signal classification accuracy of 99.99%, and a maximum average multiple recognition time of only 1.3 min.