Wheel-Rail Adhesion Sensing With AI Optical Self-Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wheel-rail adhesion measurement systems face challenges with indirect optical solutions requiring complex calibration and being susceptible to environmental conditions, and direct adhesion measurements necessitate periodic braking force application, which is unnecessary for vehicle travel.
Innovation Solution
A system using artificial intelligence for self-calibration of optical sensors based on direct adhesion measurements, allowing for reliable continuous wheel-rail adhesion measurement without the need for frequent braking by storing and comparing optical information with associated real adhesion values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If indirect optical solutions are used for adhesion measurement, then the system can operate continuously without braking, but the measurement reliability is poor due to complex calibration requirements and environmental susceptibility
Solution Approach 1:
The system performs self-calibration by automatically comparing optical sensor readings with direct adhesion measurements from brake-based measurements, eliminating the need for manual calibration procedures and enabling the system to adapt to environmental conditions autonomously
Solution Approach 2:
The system uses feedback from direct adhesion measurements to continuously refine and update the calibration of optical sensors, creating a closed-loop system that improves measurement reliability over time while maintaining continuous operation capability
2Reliability
If direct adhesion measurements are performed using dedicated brake control, then measurement reliability is improved, but the device complexity and energy consumption increase due to periodic braking force application
Solution Approach 1:
The brake system serves dual purposes: it performs its primary function of vehicle deceleration and simultaneously provides adhesion measurement capability, eliminating the need for a separate dedicated measurement brake system
Solution Approach 2:
The system automatically determines when brake-based adhesion measurements are needed and executes them autonomously, reducing the need for complex dedicated control systems while maintaining measurement reliability
3Measurement precision
If direct adhesion measurements are performed during braking, then accurate adhesion values are obtained, but productivity is reduced due to interruption of normal vehicle travel
Solution Approach 1:
The system performs brake-based adhesion measurements periodically at strategically selected moments during vehicle operation, obtaining accurate measurements while minimizing disruption to normal travel and maintaining productivity
Solution Approach 2:
The system uses optical sensors to predict adhesion conditions in advance, allowing brake-based measurements to be performed only when necessary, thereby reducing interruptions to vehicle travel while maintaining measurement precision
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A system for determining an adhesion value between wheel (W) and rail (R) for a railway vehicle (RV) is described, comprising an optical information acquisition means (3) arranged to acquire optical information from the rail (R), a database (5) arranged to store the acquired optical information in at least a learning phase and to associate said information with a relevant real wheel-rail adhesion value measured at the moment of acquisition of the optical information, a control unit (7) arranged to determine the current wheel-rail adhesion value on the basis of a comparison between current acquired optical information and the optical information previously stored in the database (5). The control unit (7) determines that the current wheel-rail adhesion value associated with the current acquired optical information corresponds to the adhesion value associated with the optical information stored in the database (5) having the highest degree of similarity with the current acquired optical information.