Overview
Automated Detection as a Solution to Rail Corrugation Noise
Rail corrugation causes high noise, passenger discomfort, and lower service quality
It shortens rail lifespan and increases derailment risk
Reliance on costly foreign inspection vehicles creates expense and staffing issues
Automated detection offers objective noise source identification and efficient track noise management.
Core Competencies
Purpose
Automated detection technology for rail corrugation ensures precise diagnosis and efficient management, overcoming the limitations of manual or costly inspection methods.
Technology acquires periodicity and amplitude from the rail surface, extracts feature vectors, and generates rail corrugation images using machine learning.
Peak frequency position of the feature vector is estimated in sub-pixel units, enabling precise calculation of corrugation periodicity.
On-site inspection of an actual operating line demonstrated diagnostic accuracy exceeding 97%.
Expected Benefits
Cost Reduction
– Replaces expensive foreign inspection vehicles, lowering overall costs.
Lower Maintenance Demand
– Reduces maintenance personnel requirements and associated expenses.
Objective Data Establishment
– Provides quantitative and objective data on rail corrugation.
Noise Issue Prevention
– Prevents railway noise problems caused by track corrugation in advance.
Expanded Diagnostic Capability
– Image-based detection technology enables diagnosis of corrugation and other major rail defects.
TRL 7-8
Contact
Korea Railroad Research Institute (KRRI)
Global Business Department
Dr. jiyoung Hong
E-mail: hongjy@krri.re.kr