Article: A novel data-driven external operational risk level distinguishing approach for catenary system of heavy-haul railway Journal: International Journal of Bio-Inspired Computation (IJBIC) 2025 Vol.26 No.2 pp.65 - 75 Abstract: The heavy-haul railway's catenary system is exposed to harsh outdoor conditions, making it vulnerable to weather-related malfunctions that can impact train safety and operations. This study presents a data-driven approach to identify the external operational risk level (EORL) of the catenary system. External risk factors are first identified, and key parameters are selected. Then, a combined method using multi-head attention (MHA), temporal convolutional network (TCN), and gated recurrent unit (GRU) is developed to assess the EORL. This approach effectively captures both short-term and long-term dependencies in time series data. The method is validated with real data from the Jingbian-Shenmu Railway, demonstrating its reliability and accuracy in distinguishing EORL for heavy-haul railways. The method is validated with real data from Jingbian-Shenmu Railway, and the proposed method achieved the best F1 score and accuracy of 99.71% and 99.45% for haze pollution and wildfire risk identification, respectively. It proves its reliability and accuracy in distinguishing heavy railroad EORL. Inderscience Publishers - linking academia, business and industry through research

Title: A novel data-driven external operational risk level distinguishing approach for catenary system of heavy-haul railway

Authors: Zhichun Zhang; Long Yan; Zhikun Wang; Tianqi Su; Tianjun Du; Lintao Wang; Miao Li

Addresses: Shaanxi Railway and Logistics Industry Group Co., Ltd, Xian 710000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China ' Shaanxi Jingshen Railway Co., Ltd, Yulin 719000, Shanxi, China

Abstract: The heavy-haul railway's catenary system is exposed to harsh outdoor conditions, making it vulnerable to weather-related malfunctions that can impact train safety and operations. This study presents a data-driven approach to identify the external operational risk level (EORL) of the catenary system. External risk factors are first identified, and key parameters are selected. Then, a combined method using multi-head attention (MHA), temporal convolutional network (TCN), and gated recurrent unit (GRU) is developed to assess the EORL. This approach effectively captures both short-term and long-term dependencies in time series data. The method is validated with real data from the Jingbian-Shenmu Railway, demonstrating its reliability and accuracy in distinguishing EORL for heavy-haul railways. The method is validated with real data from Jingbian-Shenmu Railway, and the proposed method achieved the best F1 score and accuracy of 99.71% and 99.45% for haze pollution and wildfire risk identification, respectively. It proves its reliability and accuracy in distinguishing heavy railroad EORL.

Keywords: catenary system; heavy-haul railway; weather conditions; data-driven; external operational risk level distinguishing; multi-head attention; MHA; temporal convolutional network; TCN; gated recurrent unit; GRU.

DOI: 10.1504/IJBIC.2025.149077

International Journal of Bio-Inspired Computation, 2025 Vol.26 No.2, pp.65 - 75

Received: 09 Nov 2024
Accepted: 08 Apr 2025
Published online: 13 Oct 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: A novel data-driven external operational risk level distinguishing approach for catenary system of heavy-haul railway

Keep up-to-date