Google's AI department and Harvard University researchers have developed an AI model that can predict the location of aftershocks within one year after an earthquake. This model is based on the data of 199 major earthquake disasters and 130,000 aftershocks. This model is more accurate than the methods currently used to predict aftershocks.
The main data of the model comes from famous earthquakes such as the 2004 Sumatra earthquake, the 2011 Japan earthquake, the 1989 San Francisco earthquake and the 1994 Los Angeles earthquake. After the model is completed, the earthquake and aftershock data of the last 10 years have been used for testing.
The results of this research were published in the latest edition of the journal Nature. The researchers involved in this research included Brendan Meade, professor of earth and star sciences at Harvard University, and Martin Wattenberg and Fernanda Viégas, researchers of Google's deep learning. Although these scholars are all earth-related scientists, But no seismic experts participated in the research.
AI model learning and training is used to explore the big problem of "what causes an earthquake?" Professor Meade said, "Most of the neural networks are difficult to explain, sometimes called black boxes, but for this problem, because of the understanding of physics, So know that this is an important fact conveyed through elastic pressure".
Professor Meade said that "the research results have proved to be interpretable. In fact, the cause of the earthquake is different from the physical explanation, which provides a new direction for research."
This model cannot take into account earthquakes caused by volcanic eruptions and other major natural disasters. Despite the success of this research, it is far from ready to be deployed in the real world. First, the AI ​​model only focuses on aftershocks caused by permanent ground changes, called static stress. However, subsequent earthquakes may also be caused by the ground rumble that occurred later, which is called dynamic pressure. Existing models are also too slow to work in real time.
0.8mm Pin Header
Antenk 0.8mm Pitch Male Header series is a fine pitch, low profile, single/dual/three/four row, PCB mounted connector set intended for limited space applications or where total weight is a factor. Our specially tooled insulators and contacts maintain consistent high quality through our automated production processes. Each series is available in thru-hole PCB or SMT mounting and plated tin, gold or selective gold as specified.
0.8mm Pin Header Options
Number of Rows
1/Single
2/Double
3/Three
4/Quad
Number of Positions
2 Position
3 Position
4 Position
5 Position
6 Position
8 Position
10 Position
12 Position
14 Position
15 Position
16 Position
17 Position
20 Position
Termination Style
SMD/SMT
Through Hole
Mounting Angle
Right Angle
Straight
0.8mm Pin Header Specifications:
Material: Standard Hi-Temp insulator: Nylon 6T, rated UL94V-0
Insulator Color: Black
Contacts: Phosphor Bronze
Plating:
U = Gold over nickel underplate
SG = Gold over nickel underplate on
contact area, tin over copper underplate on tails.
T = Tin over copper underplate overall.
Electrical:
Operating voltage: 250V AC max.
Current rating: 1 Amp max
Contact resistance: 20 mΩ max. initial
Insulation resistance: 5000 MΩ min.
Dielectric withstanding voltage: 1000V AC for 1 minute
Mechanical:
Mating durability: 500 cycles min.
Temperature Ratings: Operating temperature: -40°C to +105°C
Max process temp: 230°C for 30 ~ 60 seconds (260°C for 10 seconds)
Soldering process temperature: 260°C
Packaging:
Anti-ESD plastic bags or tubes
Approvals and Certifications:
UL Recognized File no. E224053
Pin Header,0.8Mm Male Header,0.8Mm Pin Header,0.8Mm Male Header Pins,0.8mm Pitch Pin Header,SMT 0.8mm Pin Header, THT 0.8mm Pin Header
ShenZhen Antenk Electronics Co,Ltd , https://www.pcbsocket.com