Repozytorium publikacji - Politechnika Gdańska

ABSTRACT Undertaking long-term acoustic measurements on sites located near an airport is related to a problem of large quantities of recorded data, which very often represents information not related to flight operations. In such areas, usually defined as zone of limited use, often other sources of noise exist, such as roads or railway lines treated is such context as acoustic background. Manual verification of such recorded data is a costly and time-consuming process. However, the use of special systems or devices that support the recognition of noise sources is an additional high cost. Thus, the said problems bring up an idea to use automated methods to identify noise sources and time of such events. The proposed solution is to use pattern recognition techniques for acoustic signal recorded by the monitoring station, without the need for additional equipment. Methods based on an automatic "meaning" of sounds recorded by measuring microphone (included monitoring station) and refer to the advanced techniques of cognitive referring to the human auditory perception. In this way will be possible to assign markers to acoustic events and calculate an acoustic indicator of environmental assessment. The recordings from the stations of long-term acoustics monitoring have been used while classifying the traffic noise - aircraft operations, movement of trains or cars.

Autorzy

• Maciej Kłaczyński,
• Witold Cioch,
• Tadeusz Wszołek,
• Wiesław Wszołek,
• Dominik Mleczko,
• Paweł Pawlik,
• mgr inż. Anna Grzeczka link otwiera się w nowej karcie