Machine Learning-Based Groundwater Level Change Prediction Method and System

By standardizing multi-source time-series data and decomposing it using a variational mode decomposition network guided by physical information, combined with collaborative training using a differentiable simulator, the problem of the lack of physical mechanisms in existing models is solved, and interpretability and robust prediction of groundwater level changes are achieved.

CN121765398BActive Publication Date: 2026-05-26INST OF KARST GEOLOGY CAGS
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Patents(China)
Current Assignee / Owner
INST OF KARST GEOLOGY CAGS
Filing Date
2025-12-26
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing purely data-driven groundwater level change prediction models lack physical mechanisms, resulting in weak interpretability and insufficient reliability of oversample extrapolation.

Method used

By standardizing multi-source time series data, a supervised learning sample set is constructed. Then, a variational mode decomposition network guided by physical information is used to decompose the groundwater level sequence into K intrinsic mode component sequences and residual term sequences. A differentiable simulator is dynamically assembled for collaborative training to generate a groundwater level change prediction sequence for future periods. Finally, destandardization and uncertainty quantification are performed.

Benefits of technology

It enhances the model's mechanistic explanatory power and extrapolation robustness in unknown scenarios, and generates reliable visual prediction results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121765398B_ABST
    Figure CN121765398B_ABST
Patent Text Reader

Abstract

This invention discloses a machine learning-based method and system for predicting groundwater level changes, relating to the field of hydrogeology. The method includes: standardizing multi-source time-series data to obtain a standardized multivariate time-series data matrix; constructing a supervised learning sample set; inputting the groundwater level sequence from the supervised learning sample set into a physically-guided variational mode decomposition network; decomposing the groundwater level sequence into K intrinsic mode component sequences and a residual term sequence using a loss function with physical-driven consistency constraints; for each of the K intrinsic mode component sequences, dynamically assembling a differentiable simulator from a library of differentiable simplified physical simulators, and co-training these simulators with the goal of approximating each intrinsic mode component sequence and reconstructing the original water level sequence as a whole, resulting in K fully trained assembled differentiable simulators. This invention generates reliable, visualized prediction results through multi-simulator collaborative extrapolation and uncertainty quantification.
Need to check novelty before this filing date? Find Prior Art