A Novel Stagger Prediction Method for Overhead Rigid Conductor Systems Using Force Measurements

Overhead Rigid Conductor Systems (ORCS) are widely used in modern urban rail networks, where precise monitoring of contact wire geometry is critical for safe operation. Traditionally, these critical parameters have been primarily obtained through expensive and environmentally sensitive industrial camera systems, presenting significant limitations. This work presents a novel framework for predicting dynamic stagger and localizating section overlaps within ORCS, offering a more cost-effective and robust alternative. The methodology integrates three components: a beam-based model to obtain dynamic stagger under moving-load conditions, a difference matrix representation with kurtosis-guided lag selection and prominence-informed peak detection for overlap localization, and zero-phase Butterworth filtering to suppress dynamic pulsations. The framework was validated on 32 distinct overlap segments across both triangular and sinusoidal ORCS geometries. The section overlap classifier achieved an $F_{\beta }$-score of 1 for both layout types, indicating 100% identification of overlaps. Furthermore, the framework exhibits excellent prediction of the stagger probability distribution, with Bhattacharyya distances between measured and predicted distributions of 0.0115 for triangular layouts and 0.0517 for sinusoidal layouts. The section-wise mean Bhattacharyya distance was validated as 0.0734, and the framework maintained robustness across ±10% speed fluctuations. This research provides a reliable, robust, and economically viable method for ORCS dynamic stagger monitoring.