Affiliation:
1. RUT(MIIT); Sirius University of Science and Technology
Abstract
Aim. The paper aims to examine the matters related to increasing the objectivity of evaluation of the quality of train control by train drivers. Methods. The study presented in the paper uses statistical analysis and linear algebra. Results. An algorithm was developed for defining preventive measures and their application efficiency was evaluated for drivers of rapid transit trains. The algorithm for defining preventive measures for drivers of rapid transit trains includes the following: violation prediction; definition of the factors that affect the onset of each type of violations; definition of the characteristics of the drivers that most deviate from the target values. The efficiency estimation is based on the assumption of correlation between the cost of a driving instructor’s work with a driver and the cost of losses that the company might incur in case of violations. The paper shows that the level of an error of the first kind in the train driver violation prediction model is justified, provided that the cost incurred as the result of gross train control violations is significantly greater than that associated with the training of such driver. The paper presents an analysis of the application of the AI-based system in four depots. Conclusion. The paper presents an algorithm for defining preventive measures for train drivers. An economic criterion was defined for evaluating the efficiency of application of the developed mathematical model for predicting gross violations of train control. The required and sufficient conditions of economic efficiency of the AI-based systems application were analysed. A comparative analysis was presented of the mean number of gross train driving violations in depots with and without the AI-based system.
Reference14 articles.
1. Baranov L.A., Sidorenko V.G. [Smart cyberphysical rapid transit control systems (Part 2)]. Transportnoye stroitelstvo 2021;4:19-23. (in Russ.)
2. Baranov L.A., Sidorenko V.G., Balakina E.P., Loginova L. N. Intelligent centralized traffic management of a rapid transit system under heavy traffic. Dependability 2021;21(2):17-23. DOI 10.21683/1729-2646-2021-21-2-17-23.
3. Kulagin M.A., Kharin O.V., Yakimov S.M., Gonik M.M., Khludeev M.A., Yaroshchuk D.I. [Automated System Trusted Environment of the Locomotive Service. Russian state registration certificate for software no. 2020612670]. Published on 23.03.2020.
4. Kulagin M., Sidorenko V. Recommender Subsystem Construction for Calculating the Probability of a Violation by a Locomotive Driver using Machine-learning Algorithms. IEEE East-West Design & Test Symposium (EWDTS); 2020. P. 1-5.
5. Kulagin M., Sidorenko V. An Intelligent Evaluation System for Predicting Engine Driver Reliability. Russian Electrical Engineering 2020;91(9):587-591.