Statistical assessment of environmental noise generated by road traffic / Statistinis aplinkos triukšmo, sklindančio nuo kelių transporto, įvertinimas / Статистическая оценка шума, производимого автомобильным транспортом

Transport

Marius Mickaitis¹  Boris Fiks²  Aleksandras Jagniatinskis² 

[1] Dept of Building Structures, Vilnius Gediminas Technical University, Pylimo g. 26/1, LT-01132 Vilnius, Lithuania;Institute of Thermal Insulation, Vilnius Gediminas Technical University, Linkmenų g. 28, LT-08217 Vilnius, Lithuania

关键词: acoustics;    environmental noise;    urban traffic noise;    statistical analysis;    histograms;    percentiles;    measurements;    short-term assessment;   

DOI  :  10.3846/16484142.2011.568084

学科分类：航空航天科学

来源: Vilnius Gedinimas Technical University

【 摘 要 】

The necessity to check requirements for sound insulation of new building façades against outdoor noise and for estimating noise pollution of the residential area comprise the assessment of environmental noise taking in situ measurements that are mainly carried out under the standardized method presented in the EU directives and Lithuanian legislation. For these estimations, annual composite daily equivalent sound pressure level Ldеn becomes obligatory. When the value of this descriptor is determined from measurements, it is relevant to reasonably reduce a time interval of the performed measurement. The possibilities of determining the noise level Ld ,12h of the day (12h) undertaking annual Lden assessment making short-term (15÷60 min.) measurements is dealt in this work. In this way, the sensitivity of the set of statistical descriptors to variations in sound pressure levels was investigated. The paper proposes that the obtained instantaneous A-weighted sound pressure levels with F (fast) time weighting histogram of distribution density may be used for characterizing noise pollution in the sites. When traffic noise emitted from arterial town roads with stationary traffic takes a dominating position, the form of these histograms is close to normal distribution density with corresponding statistical parameters: average, standard deviation, asymmetry coefficient and percentiles. The experimental measurements have established that an asymmetry coefficient and an introduced descriptor– the rank of percentile Neq% are the most sensitive to noise variations. The last one is more preferable due to its exclusive relationships with an equivalent level– the corresponding value of this rank of percentile PNeq that coincides with the equivalent level of the measured time interval.

【 授权许可】

CC BY