Development of an algorithm for constructing a wheel profile and detecting surface defects during wheel-rail interaction

The global growth of rail transportation contributes to economic development, but at the same time increases the safety requirements for rolling stock. Wheelset defects, which are the cause of most accidents, remain a key risk factor. To detect them in a timely manner, a comprehensive analysis of the wheel surface is required, including image collection, restoration of damaged areas and accurate measurement of geometric parameters. Optical systems are used for timely detection of defects, which allow for real-time monitoring of the surface. However, the effectiveness of such systems largely depends on the algorithms used to calculate and evaluate the parameters. The presented paper considers the relevance of developing an improved data processing algorithm for monitoring the main parameters of the wheels in order to improve the accuracy of diagnosing rolling surface defects.