A gold mine target area prediction method and system based on ore-forming causal orientation

By introducing the PCMCI algorithm and geological prior constraints to construct a mineralization causal map and extracting causal guidance features, the problem of low accuracy and poor interpretability of gold ore target area prediction caused by the neglect of causal mechanisms in existing technologies is solved, and efficient and accurate target area prediction is achieved.

CN122064969BActive Publication Date: 2026-07-07XIAN CENT OF GEOLOGICAL SURVEY CGS +1
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
XIAN CENT OF GEOLOGICAL SURVEY CGS
Filing Date
2026-04-03
Publication Date
2026-07-07

AI Technical Summary

Technical Problem

Existing gold target prediction technologies ignore the causal mechanism between variables, rely on correlation and are easily affected by noise, and lack the integration of geological prior knowledge, resulting in low prediction accuracy and weak interpretability in shallow overburden areas.

Method used

The PCMCI algorithm, combined with geological prior constraints, is used to construct a mineralization causal map and extract causal guidance features. The model hyperparameters are then optimized using the gradient descent algorithm to predict gold ore target areas.

Benefits of technology

It improves the robustness and interpretability of the model, significantly enhances the accuracy of target area prediction in shallow-covered areas, and provides scientific evidence to support exploration and deployment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122064969B_ABST
    Figure CN122064969B_ABST
Patent Text Reader

Abstract

The application relates to the technical field of intelligent exploration, in particular to a gold mine target area prediction method and system based on metallogenic causality orientation, which is applied to gold mine target area prediction in a shallow coverage area and comprises the following steps: obtaining geophysical and geochemical data of a region to be predicted, inputting the geophysical and geochemical data into a target area prediction model, and obtaining a prediction result of whether the region to be predicted is a target area; wherein the target area prediction model is trained through the following steps: obtaining historical data of a target area; obtaining a historical metallogenic causality graph through a PCMCI algorithm based on a geological prior constraint set; performing causality feature extraction on the historical metallogenic causality graph to obtain corresponding causality orientation features; inputting the causality orientation features into the target area prediction model, optimizing model hyperparameters through a gradient descent algorithm, and obtaining a target area prediction model for predicting a gold mine target area in a shallow coverage area. The model has improved robustness and interpretability, and precise target area prediction in a shallow coverage area is realized.
Need to check novelty before this filing date? Find Prior Art