Spectral Analysis of Forest Fire Noise for Early Detection using Wireless Sensor Networks
| Parent link: | Control and Communications (SIBCON): Proceedings of the XII International Siberian Conference, Moscow, May 12–14, 2016. [4 p.].— , 2016 |
|---|---|
| Main Author: | |
| Corporate Author: | |
| Other Authors: | |
| Summary: | Title screen Crown fires are extremely dangerous, very difficultto fight and often have a rate of spread over 100 times more thana surface fire. Therefore, it is important to determine the type offorest fire in the early detection based on wireless sensor networks (WSNs) to adopt the proper strategy to fight the fire. It is shownthat this could be done analyzing the noise power spectrum offorest fires: surface fires noise spectrum can be modeled as the rednoise (gradual increase of trend line amplitude toward lowerfrequencies), while for crown fires noise spectrum trend line hasan almost bell-shaped (Gaussian) type. The noise frequency rangeis relatively narrow for crown fires and ranged from 250 to 450Hz. The intermediate type of fires (strong surface fire andincipient crown fire) has a transient noise spectrum frombroadband red to narrowband Gaussian. The article presents thespectrums of 9 different forest fires. The different trend line of theforest fire noise power spectrum is the parameter that can be usedto determine the type of forest fire in WSNs. |
| Published: |
2016
|
| Series: | Sensors and Systems of the Internet of Things |
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1109/SIBCON.2016.7491654 |
| Format: | Electronic Book Chapter |
| KOHA link: | https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=648806 |