China University of Science and Technology team pioneers artificial intelligence full-automatic seismic monitoring system to report source parameters within 2 seconds

Hefei, 26 February (Reporter Wu Lan) Reporter learned from the University of Science and Technology of China on the 26th that after six years of hard work, several research groups of the school and Zhao Cuiping, a research team of the Institute of Earthquake Prediction, China Earthquake Administration, launched the world The first artificial intelligence seismic monitoring system-the "smart ground motion" monitoring system. The system has been used for one year in the China Earthquake Experimental Field, and all seismic source parameters are reported in 1-2 seconds.

At present, the system's R & D team is discussing exchange, cooperation, and promotion plans with Japan, Turkey, Mexico, and other national earthquake monitoring agencies.

When an earthquake occurs, the seismic wave signal is transmitted to the seismic network. Seismic professionals use theoretical algorithms to process seismic data, and obtain the parameters such as the epicenter, magnitude, time of occurrence, depth, and rupture mechanism of the earthquake. Disaster reduction provides important scientific information.

Although many countries use automatic fast digital algorithms to obtain seismic information, the results are often unreliable and false positives often occur. Therefore, the combination of automatic algorithms with manual calculation results and quality monitoring by professionals is the main means of monitoring earthquakes at present. The method requires that the seismic monitoring organization must strengthen the ability of manual continuous monitoring on a 24-hour basis.

After six years of hard work, Professor Zhang Jie's research group at the School of Earth and Space Sciences, University of Science and Technology of China, Professor Li Weiping, Professor Chen Zhibo's group at the School of Information Science and Technology, Professor Chen Enhong's group at the School of Computer Science and Technology, and China Earthquake Administration's earthquake prediction research Researcher Zhao Cuiping's team collaborated to launch the world's first artificial intelligence seismic monitoring system, the "Intelligent Ground Motion" monitoring system.

The system has been used for a year at the China Earthquake Experimental Fields in Sichuan and Yunnan provinces. Based on the comparison of 446 earthquake evaluation results in 2019, the China Earthquake Administration showed that the unmanned "smart ground motion" monitoring system and manual calculations were performed by professionals. The accuracy is very close. At the same time, the "smart ground motion" monitoring system only needs 1-2 seconds to report all the seismic source parameters, while the world's most advanced automatic rapid reporting system of the United States Geological Survey (USGS) requires 3 -Earthquake information is reported after 5 minutes of calculation time.

According to reports, the ability to quickly report earthquake early warning information is of great significance to promptly notify the public to take immediate protective measures, quickly terminate the operation of important facilities such as nuclear power plants, slow down high-speed rail, and determine rescue plans. The emergence of an unmanned artificial intelligence real-time seismic monitoring system has greatly eased the pressure of 24 hours of manual monitoring of earthquakes, especially the ability to process intensive seismic network big data in time, helping to reduce false positives and false negatives.

Prof. Zhang Jie's research group and Prof. Chen Enhong's research group published the innovative work of the seismic search engine in Nature Communications in 2014. The research results of this work also realized the actual operation in the "smart motion" system and played an important role. On February 6, 2020, Professor Zhang Jie's group and Professor Li Weiping and Chen Zhibo's group jointly published related papers in ScientificReports. In the past two years, Professor Zhang Jie's group has published 15 academic papers in the field of artificial intelligence geophysics and made relevant academic reports 21 times at international academic conferences.

At present, many well-known international universities and research organizations are studying the problem of artificial intelligence seismic monitoring, but most of them are at the basic scientific research and publication stage. The cross-cooperation team of the University of Science and Technology of China cooperates with the China Earthquake Administration to launch a practically running artificial intelligence system, which is an international leader. The system has successfully tested seismic data in multiple countries.