A deep learning-based high-speed railway subgrade vibration anomaly grading early warning method

Abstract

This invention relates to the field of intelligent monitoring technology for rail transit safety, specifically to a method for graded early warning of vibration anomalies in high-speed railway subgrades based on deep learning. This invention deeply integrates subgrade dynamics priors with deep learning. Specifically, it extracts multi-frequency anomaly features conforming to physical laws through multi-scale sparse convolution, employs an adaptive gating mechanism to remove interference from operating conditions such as train speed and axle load, utilizes dual-domain attention to focus on key anomaly periods and channels, and simultaneously outputs a risk index and reliability coefficient. Combined with dynamic three-level thresholds and an adaptive sliding window, it achieves accurate graded early warning. This invention can improve the accuracy, stability, and decision reliability of high-speed railway subgrade vibration anomaly identification.

Images