RU: По результатам натурного исследования путем применения регрессионного анализа и методов математической статистики выявлены формулы для наиболее точного определения уровня звука. Предложены четыре модели для определения шумовых характеристик движущегося современного транспортного потока, произведен сравнительный анализ предлагаемых формул и базовой, выбран наиболее оптимальный вариант модели.
EN: Problem statement. Right calculation of the equivalent loudness level caused by moving vehicle flow is a serious problem nowadays. It is a questionable task because the most recent investigations in this field of science were held thirty years ago and they need to be modernized. Thereby the new model of equivalent loudness level is required nowadays.
Analyzing of the resent research. Origins of a mathematical theory of traffic flow date back to the 1920s, when Frank Knight first produced an analysis of traffic equilibrium, which was refined into Wardrop's first and second principles of equilibrium in 1952.
A lot of scientists all over the world such as R. Kartabaev, V. Syl'yanov, F. Kheyta, D. Dryu et al. were working on the problem of the equivalent loudness level calculation. However, their research was being conducted long ago. In the modern world the new model for the traffic noise calculation is required.
Research objective. The aim of the article is to suggest 4 calculation models for traffic noise calculation and compare their results with the outdate base formula.
Conclusions. According to the results obtained during the research and using mathematical statistics methods including regression analysis calculation models were detected. There are 4 models for modern traffic flow noise characteristics accurate measurement. The comparative method of inquiry is performed between 4 suggested models and the base one. The most optimal model was chosen. Calculated results make evident that the new model describes the process of noise diffusion more clearly and accurate. Calculation mean error while using the base formula is 3,49 dBA, while the new one (model 4) is 0,91 dBA. Model 4 insignificantly overrates the real noise value (0,00063 %). However, comparing with the overrating while using the base formula (3,7 %), the error is negligible.
