Creating noise pollution maps based on user-generated noise measurements

نویسنده

  • Axel Schulz
چکیده

Environmental pollution has become a rising concern during the last decades. Especially in big cities with their huge traffic volume, the pollution by the generated noise has become an imminent health threat to the citizen [40]. To protect their citizens, states in the EU have agreed on a directive which forces them to curtail the threat that is noise pollution [31]. First and foremost, this gives citizens the right to be informed about noise threats as every state has to provide noise maps for area of high population density. However, there are three major problems with this approach. Primarily, those maps only provide sparse coverage since rural areas are not included and they also exhibit a relatively long update cycle. Last but not least, recording the maps involves expensive high quality sensors and human resources, resulting in a not negligible financial burden. In this thesis, we introduce an approach making use of machine learning techniques and collected noise measurements by a participatory sensing application to determine the noisiness factor of an area. By using the collected sound data in conjunction with further information on the vicinity such as nearby streets or buildings, we were able to create a machine learning model able to predict a sound level with 80.9% accuracy. As this approach is very cost-efficient it can easily be used as an addition to common techniques for recording noise maps as well as a standalone application. Zusammenfassung Das Thema Umweltverschmutzung ist in den letzten Jahren zu einem ernstzunehmenden Problem herangewachsen. Vor allem in Städten mit großem Verkehrsaufkommen ist die Belästigung durch den hervorgerufen Lärm zu einer Bedrohung der Gesundheit der Einwohner geworden [40]. Zum Schutze ihrer Bevölkerung haben die Mitgliedsstaaten der EU eine Richtlinie verabschiedet, welche die zunehmende Lärmbelästigung eindämmen soll [31]. Durch Lärmkarten von Ballungszentren sollen Bürger über Lärmquellen informiert werden. Dieser Ansatz leidet jedoch unter drei grundsätzlichen Schwächen. Zum einen gibt es für ländliche Gebiete keine Lärmkarten und des Weiteren verhindert der meist lange Aktualisierungszeitraum eine genaue Bestimmung der Stärke der Lärmbelästigung. Zudem werden teure Qualitätssensoren sowie Fachkräfte zur Aufnahme einer Lärmkarte benötigt, was eine nicht vernachlässigbare finanzielle Belastung darstellt. In dieser Arbeit stellen wir einen Ansatz vor, welcher mithilfe gesammelter Lautstärkemessungen durch eine “Participatory Sensing” Anwendung den Lärmfaktor einer Gegend bestimmen kann. Hierzu verwenden wir Techniken des Maschinellen Lernens und reichern unsere Messungen mit einer genauen Beschreibung der Umgebung des Messorts an, beispielsweise indem wir Straßen und Gebäudetypen in der Nähe ergänzen. Dadurch konnten wir ein Modell erstellen, welches den Lärmpegel mit einer Genauigkeit von 80.9% vorhersagen kann. Da unser Vorgehen sehr kostengünstig ist, kann es leicht als Ergänzung zu herkömmlichen Methoden zur Erstellung von Lärmkarten oder als eigenständige Applikation eingesetzt werden.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Study of Noise Map and its Features in an Indoor Work Environment through GIS-Based Software

Background: Noise mapping in industry can be useful to assess the risks of harmful noise, or to monitor noise in machine rooms. Using GIS -based software for plot of noise maps in an indoor noisy work environment can be helpful for occupational hygienists to monitor noise pollution. Methods: This study was carried out in noisy packaging unit of a food industry in Ghazvin industrial zone, to ev...

متن کامل

Assessment and Control of Noise Pollution in Seal Gas Compressor of Reduction Unit 2 of a Steel Industry

Background and aims: Undoubtedly, noise and vibration are major problems of industrial world and lot of people are very exposed to these phenomena in their workplace or residential areas. Researchers have shown that exposure to noise may increase the risks related to personal health, like nervous frailty, extreme irritability, muscle cramps, stress and anxiety, dizziness, headache and migraine,...

متن کامل

Noise annoyance through railway traffic - a case study

This paper describes an assessment of noise caused by railway traffic in a large Latin American city. Measurements were taken of noise levels generated by trains passing through residential neighborhoods with and without blowing their horns. Noise maps were also calculated showing noise pollution generated by the train traffic. In addition - annoyance of the residents - affected by railway nois...

متن کامل

Evaluation of noise pollution levels before and after corrective actions in the operational units of tanks and petroleum products transportation in Abadan Oil refining Company

Introduction & Objective: Noise exposure, risk exposure in a wide range of work station, especially oil and gas industry and related companies is considered. This study was carried out to evaluate the noise before and after corrective action in 11 operational units of tanks and transportation in Abadan oil refining company. Methods: First, the basic information including the location of noise ...

متن کامل

Road Traffic Noise: Gis Tools for Noise Mapping and a Case Study for Skåne Region

Traffic noise pollution is a growing problem that highly affects the health of people. To cope with this problem one has to regulate traffic or construct noise barriers. In order to implement effective measures against traffic noise the information about its distribution – noise maps is imperative. This paper presents our work in creating a noise calculator software package implementation that ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2013