Аналіз та вибір алгоритму розпізнавання сходу вагонів швидкісного руху

Abstract

UA: У статті запропоновано алгоритм розпізнавання сходу вагонів швидкісного руху, зроблено вибір найбільш раціонального коду, що відповідає кожному стану об'єкта, запропоновано основні ознаки сходу пасажирського вагона швидкісного руху, алгоритми логічної обробки сформованих ознак, які забезпечують оптимальне рішення задачі виявлення сходу вагона, Для порівняльної оцінки якості ознак розпізнавання встановлено критерії якості і вибране оптимальне рівняння функції, що розділяє.

EN: 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.