An efficient method of forest fire detection using wireless sensor network with yagers modified dempsters rule
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Forests play a crucial role for sustaining the human environment
newlineand forest fires are among the largest dangers having physical, biological,
newlineecological and environmental consequences. Hence, it is necessary to
newlinepredict and detect forest fires more promptly and accurately. In this
newlinethesis, an intelligent system is proposed to predict and detect forest fire
newlineusing Wireless Sensor Network, as incredible advances in Wireless
newlinecommunications technology and low cost sensors have made it possible
newlineto build large scale Wireless Sensor Network (WSN) to detect forest fire.
newlineThis intelligent system consists of three different classifiers using
newlineSupport Vector Machine (SVM) for classifying the large forest data. The
newlineendeavour of using more than a single classifier is that classification in
newlinethis forest fire context is extremely important as the cost of
newlinemisclassification using a single classifier is very high. The classified data
newlineare then combined using a combination technique proposed by Yager a
newlinemodification of Dempster s rule which is a prominent frame work to
newlinehandle the uncertainty in the Network. In this thesis, Yager s modified
newlineDempster s rule, a data fusion technique is introduced along with
newlinesubsequent improvements for combining conflicting information in
newlineYager s structure. In the forest, chances of occurrence of natural hazards
newlinelike forest fires and gas outbursts are although rare but with high
newlineuncertainty thus provides weak information about the system.. To
newlinewrestle with such problems, Yager s modified Dempster s formalism is a
newlinestrong and effective tool. Here in this thesis, complete modus operandi of
newlinethe Yager s modified Dempster s rule has been narrated with the help of
newlineillustrative examples and by using this technique, it is possible to detect
newlineforest fire more promptly.
newline