A fire detection system based on ART-2 neuro-fuzzy network

Author

Qing, Zhang ; Shu, Wang

Author_Institution

Dept. of Electron. & Inf. Eng., Huazhong Univ. of Sci. & Technol., China

Volume

2

fYear

1998

fDate

1998

Firstpage

1355

Abstract

The ART-2 neural network is a self-organized artificial network that operates according to adaptive resonance theory. A neuro-fuzzy network, which combines ART-2 and the fuzzy system in series, is presented and applied to fire detection. The results of experiments show that this system has a stronger ability to adapt to the environment than the backpropagation (BP) neural network. It can detect various standard test fires more rapidly and accurately, and has strong anti-interference capability

Keywords

ART neural nets; alarm systems; fires; fuzzy neural nets; self-organising feature maps; unsupervised learning; ART-2 neuro-fuzzy network; adaptive resonance theory; anti-interference capability; fire detection system; self-organized artificial network; self-steady learning; signal preprocessing; standard test fires; unsupervised competition; Data mining; Data preprocessing; Fires; Fuzzy neural networks; Fuzzy systems; Nonlinear optics; Optical sensors; Sensor phenomena and characterization; Sensor systems; Temperature sensors;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-4325-5

Type

conf

DOI

10.1109/ICOSP.1998.770871

Filename

770871