Earthquake prediction method, device, medium and equipment based on earthquake precursor information

By extracting features and correcting attenuation of atmospheric tidal gravity anomaly signals, and combining gradient boosting models with kernel density estimation, the problem of far-field overestimation and near-field underestimation of atmospheric tidal gravity anomaly signals in strong earthquake prediction in existing technologies has been solved. This enables quantitative prediction of strong earthquake magnitude and focal range, improving the accuracy and interpretability of predictions.

CN121454592BActive Publication Date: 2026-06-05XI AN JIAOTONG UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511609949.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-11-05
Publication Date
2026-06-05
Estimated Expiration
2045-11-05

AI Technical Summary

Technical Problem

Existing technologies for predicting strong earthquakes using atmospheric tidal gravity anomaly signals lack systematic quantitative methods and suffer from problems such as overestimation in the far field and underestimation in the near field. They also struggle to achieve a stable mapping between magnitude and source distance and lack a verifiable and transferable quantitative prediction system.

Method used

Feature extraction and screening are performed using atmospheric tidal gravity anomaly signals. A decay function is introduced for physical normalization. A joint prediction framework is constructed by combining a gradient boosting model and kernel density estimation. Through spatial partitioning simulation and decay correction, quantitative prediction of strong earthquake magnitude and focal range is achieved.

Benefits of technology

It improves the accuracy of strong earthquake prediction, provides spatial information on magnitude and high-risk locations, enhances the interpretability and stability of prediction results, and can provide quantitative prediction results several days to twenty days before an earthquake, supporting earthquake risk assessment and emergency response.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121454592B_ABST
    Figure CN121454592B_ABST
Patent Text Reader

Abstract

The application discloses a strong earthquake prediction method and device based on earthquake precursor information, a medium and equipment, and relates to the technical field of earthquake prediction. When the atmospheric tidal gravity signal is determined to be an atmospheric tidal gravity anomaly signal, the atmospheric tidal gravity anomaly signal is input to a gradient boosting machine model constructed in advance, the research area is simulated by spatial partitioning, and the simulated magnitude of each partition is obtained. Then, the energy loss of the signal in the complex medium propagation is compensated by using an attenuation function, and the actual magnitude of each partition is obtained. Further, in combination with the spatial distribution characteristics of regional earthquakes, the distribution of the source risk area is determined, and the strong earthquake prediction result of the research area is output accordingly. On this basis, by introducing a feature contribution degree analysis method, the physical interpretability of the model output is tested, and the influence mechanism of the key features of the abnormal signal on the prediction result is determined. The method improves the accuracy of quantitative prediction of strong earthquakes.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method for building seismic oscillation attenuation relation based on two-stage residual analysis

    CN103926621A

  • Large earthquake forecasting method based on weak transient disturbance signal

    CN118534557A