Russia has proposed a way to improve safety on railways

MOSCOW, November 29Siberian Federal University scientists have created algorithms and software for predicting lateral wear of rails on railways. According to the authors, their development can improve safety on curved sections of railways, as well as reduce labor costs for predictive calculations and preparation of documentation based on the results of forecasting rail wear. The research results were presented at the 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering.
Researchers at Siberian Federal University (SFU) explained that currently, prediction of lateral rail wear is carried out manually by the operator, without the use of special tools. Calculations are made using standard office software in tables, the initial data for which are statements with the results of measuring the lateral wear of rails by a track meter for the period preceding the forecast.

«

“When trains move along curved sections of track with a radius of no more than 1000 meters, lateral wear of the rails occurs. If you do not promptly monitor the excess of permissible wear values, you can cause accidents of varying severity; in the worst case, the train can go off the track. To prevent such incidents on railways, it is necessary not only to measure the current values ​​of lateral wear of rails in curves, but to predict them in order to work proactively,” explained the relevance of the development, Associate Professor of the Department of Business Informatics and Business Modeling processes of SFU Roman Kuzmich.

Scientists have developed a comprehensive solution to the difficulties of forecasting. An algorithm for predicting lateral wear of rails on curved sections of railways was developed and implemented in the form of software.

The resulting software has the ability to process a large number of initial data (more than 10,000 records), allows you to calculate predictive wear values ​​for the required number of periods, and also performs all calculations and generates the resulting information automatically.
The program also issues recommendations based on the results of the lateral wear forecast on the maximum permissible speeds of railway transport on rails with lateral wear in accordance with the relevant instructions on replacing such rails as a matter of priority.

When developing algorithmic software, a synthesis of approaches was used , which are the basis of the methodology for determining the intensity of lateral wear of rails and mathematical methods of forecasting (regression analysis models).

In the future, a map of a section of the railway track can be linked to the developed software, on which the results of forecasting lateral wear will be displayed rails in curves. Automatic distribution of output documentation to specified corporate email addresses of railway company employees can also be implemented.
The research was carried out with the support of the Ministry of Science and Higher Education of the Russian Federation under grant No. 075-15-2022-1121.