A method for predicting great earthquake magnitude by HR-GNSS combining physical perception and attention mechanism

CN121899887BActive Publication Date: 2026-05-29SHANDONG UNIV OF SCI & TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHANDONG UNIV OF SCI & TECH
Filing Date
2026-03-20
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing magnitude estimation methods based on HR-GNSS and deep learning struggle to adaptively differentiate between high signal-to-noise ratio and low-quality stations under multi-station input conditions. Furthermore, they fail to adequately incorporate the geometric relationship between the seismic source and the station, resulting in insufficient stability and generalization ability of the model in different regions and complex observation environments.

Method used

A large earthquake magnitude prediction method based on HR-GNSS that integrates physical perception and attention mechanisms is adopted. The method extracts waveform features from multiple stations through a convolutional neural network, introduces a station-dimensional dot product self-attention mechanism for feature interaction, and combines physical auxiliary features of epicentral distance for feature fusion to output the magnitude prediction value.

Benefits of technology

It enables rapid and accurate estimation of large moment magnitudes under complex observation conditions, improves the robustness and stability of the model, can adaptively identify the characteristics of high-quality stations, suppress interference from low signal-to-noise ratio stations, and enhances the model's adaptability in different regions and observation environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121899887B_ABST
    Figure CN121899887B_ABST
Patent Text Reader

Abstract

The application belongs to the technical field of earthquake magnitude determination, and specifically discloses a HR-GNSS large earthquake magnitude prediction method fusing physical perception and attention mechanism. The method first acquires HR-GNSS three-component waveform timing data of multiple stations, extracts timing characteristics through multi-layer convolution, and compresses along the time dimension to form a feature matrix corresponding to multiple stations. Then, a dot product self-attention mechanism is introduced in the station dimension to build the correlation between different station features and perform information interaction, and a global waveform feature vector representing an earthquake event is output. The magnitude prediction model also takes the epicentral distance of each station as auxiliary input, extracts a geometric distance feature vector of the spatial distribution of the station from the epicentral distance vector through a multi-layer perception machine. Finally, the obtained global waveform feature vector and geometric distance feature vector are fused in a high-dimensional space, and a prediction result of the magnitude of an earthquake is output through a fully connected regression module. The method can realize fast and accurate estimation of the magnitude of a strong earthquake.
Need to check novelty before this filing date? Find Prior Art