Determination model of the apparatus state for railway automatics with restrictive statistical data

Abstract

In the report, microprocessor systems of railway automation are represented by a distributed system of computing [1]. Under certain conditions it is expedient to represent them as a non-oriented graph and optimize with the use of the method of the least-clique [2]. The prognostic model and the method of determining the failures of hardware of microprocessor systems of railway automatics have been developed. They allow you to determine the probability of a device failure from a particular group using Student’s t-distribution, maximum likelihood estimation and uneven observations. Unlike existing approaches to forecasting, the proposed method takes into account the limited amount of control systems operation statistical data.