МЕТОДИКА ПРОГНОЗИРОВАНИЯ ГЕОФИЗИЧЕСКИХ СОБЫТИЙ В ЗЕМЛЯНОМ ПОЛОТНЕ ЖЕЛЕЗНОЙ ДОРОГИ

Abstract

The relevance of this topic stems from the need to develop accurate algorithms for predicting geophysical events and monitoring the state of the railway subgrade. This will enable timely notification of potential hazards and the implementation of measures to minimize damage. During the operation of the railway track, the integrity of the earthen base can be compromised due to exogenous processes such as karst phenomena, suffusion, and other geodynamic processes that cause deformations and reduce the bearing capacity of the soil base. This paper examines the use of machine learning algorithms based on neural networks for accurately predicting seismic events and detecting changes in the subgrade by analyzing changes in the phase signal recorded during electrophysical measurements. Particular attention is paid to the analysis of phase shifts in geoelectric signals and their use for a detailed study of the soil structure and the identification of hidden defects. Models have been developed that allow for the assessment of the integrity of the earthen base by observing changes in the phase signal. The results obtained confirm the potential of using intelligent data processing and phase signal analysis methods in geophysical monitoring and forecasting, which contributes to improving the safety of railway infrastructure.