System for managing the technical condition of a locomotive fleet on the basis of an artificial neural forecasting network

The goal of the research is to development of a synchronous-replicated model for the assessment of the technical state of a locomotive as a technical system to reduce the occurrence of failures during operation, and as a result, reduce downtime in repairs. When performing the research, the following interdisciplinary and mathematical methods were used: system analysis, computer and mathematical modeling, methods of the theory of artificial neural networks, mathematical analysis. As a result of the research, a mathematical synchronously replicated model for assessing the technical condition of a locomotive based on an artificial multilayer forecasting neural network was obtained. The developed model can be used in monitoring systems, control, diagnosing the technical condition of the locomotive fleet. The original features of the developed model are a low sampling period between polling monitoring tools, versatility, adaptability, efficiency. Based on the developed model, a generalized algorithm for managing the technical condition of the locomotive fleet is built. The proposed model and algorithm solves the ranges of tasks described in the development concept of Russian Railways OJSC related to the implementation of the actual repair system according to the current technical condition of the locomotive, as well as the digitalization of the company’s advanced areas.