ANALYSIS AND SELECTION FOR THE ALGORITHM IDENTIFICATION OF THE DERAILMENT OF HIGH-SPEED CARRIAGES

Abstract

The purpose of this work is to develop algorithms for recognizing the gathering of high-speed carriages. For this, the problem of choosing the most rational code corresponding to each state of the object was solved, the optimal equation of the separating function was chosen. For this, the problem of choosing the most rational code corresponding to each state of the object was solved, the optimal equation of the separating function was chosen. In general, the detection of a derailment of a railcar consists of the following operations: the conversion of monitored parameters into digital form for further processing by means of computer equipment; formation of signs (code) of the state of the object; classification of the state of the object in accordance with certain requirements. To solve these problems, it is necessary to study the design features of the running gears and the superstructure, their working conditions, signs (parameters) characterizing the descent of the car, the behavior of these parameters in different modes of train movement. Also, the parameters of the recognition system hardware (sensor characteristics, microcontroller capabilities, power supplies for electronic circuits) must be taken into account. The main controlled parameters in this process are the amplitude and frequency of oscillation of the wheelset. Here, the system is required to separate the values of these parameters to identify a descent - these are road irregularities, wheel defects. For a comparative assessment of the quality of recognition features, it was proposed to establish quality criteria. Here a criterion is used, based on a comparison of the statistical characteristics of the signs. The classification of the state of an object is as follows: the likelihood ratio is calculated when the feature vector is measured and a hypothesis is then accepted or rejected depending on whether the point found is located above or below the separating function. The decision making algorithm in favor of one of the classes of states of recognition objects is recommended to use the decision method with a constant sample size. The separation function equation is derived from the Bayesian criterion, which minimizes the average risk of making a wrong decision.

Authors and Affiliations

Tetyana Vojnarovs`ka, Nadiya Dyeordiyeva, Svitlana Saulyak

Keywords

Related Articles

MODERN APPROACH TO RESEARCH OF BIOGENIC CORROSION OF SEWAGE COLLECTOR

The work is devoted to the study of biogenic corrosion of sewer collectors. The analysis of newest works, approaches, practices, devoted to the study of corrosion processes of sewerage collectors has been carried out. Th...

GEOINFORMATION OF INTEGRATED ASSESSMENT OF SPATIAL FACTORS OF LAND OF MEGAPOLIS

We present the directions of the definition of geoinformation support for an integrated assessment of the spatial factors of land in a megacity. Systematized theoretical provisions for the definition of geoinformation su...

FORMING OF SPATIAL INFLEXIBILITY OF RAIL SUPPORTS AT RAILROAD TIES SB-3-0 AND FASTENING КPP-5

Аt action on the rail of forces vertical and lateral there is its bend in a plane vertical and horizontal and its twisting. The reactions of the rail fastening and reactive croutni vertical and horizontal are the constit...

DEFINITIONS URBAN DEVELOPMENT FACTORS THAT AFFECT THE USE OF CITY LANDS

We represent the main areas and urban indicators determining factors that influence land use cities. Designated urban factors can build a system of land relations cities. Of particular importance are identified stakehold...

ENGINEERING METHODS FOR CALCULATION OF REINFORCED CONCRETE STATICALLY UNDEFINED ROD SYSTEMS

Methods are proposed for calculating statically indeterminate rod systems, taking into account the nonlinear work of materials and taking into account cracks. The method of discrete registration of the stiffness of secti...

Download PDF file
  • EP ID EP524435
  • DOI -
  • Views 47
  • Downloads 0

How To Cite

Tetyana Vojnarovs`ka, Nadiya Dyeordiyeva, Svitlana Saulyak (2018). ANALYSIS AND SELECTION FOR THE ALGORITHM IDENTIFICATION OF THE DERAILMENT OF HIGH-SPEED CARRIAGES. Збірник наукових праць Українського державного університету залізничного транспорту, 180(14), 39-45. https://europub.co.uk/articles/-A-524435