Forest fire detection: A fuzzy system approach based on overlap indices
| Autor: | Humberto Bustince, Santiago Garcia-Jimenez, Miguel Pagola, Aranzazu Jurio, Laura De Miguel, Edurne Barrenechea |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
0209 industrial biotechnology
Adaptive neuro fuzzy inference system Generalization Fire detection business.industry Inference 02 engineering and technology Machine learning computer.software_genre Fuzzy logic Expression (mathematics) 020901 industrial engineering & automation 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Convex combination Data mining Artificial intelligence business computer Wireless sensor network Software Mathematics |
| Zdroj: | Applied Soft Computing. 52:834-842 |
| ISSN: | 1568-4946 |
| DOI: | 10.1016/j.asoc.2016.09.041 |
| Popis: | Graphical abstractDisplay Omitted HighlightsWe propose a new inference algorithm using overlap functions and indices.The convex combination of overlap expressions maintains the overlap properties.We avoid the difficult selection of appropriate expressions for each problem.We use our new inference algorithm in a Fuzzy Logic System to detect forest fire. It is well known that a powerful method to tackle diverse problems with lack of knowledge and/or uncertainty are Fuzzy Logic Systems (FLSs). In the literature, there exist different fuzzy inference mechanisms based on fuzzy variables and fuzzy rules to obtain a solution. In this work we introduce a generalization of the inference algorithm proposed by Mamdani, by using overlap functions and overlap indices. A challenging issue is the selection of most suitable overlap expressions for each problem. For this aim, we propose to use the convex combination of several ones. In this way, the conclusions obtained by our FLSs avoid the bad results obtained by an inadequate overlap expression. We test our proposal on a real problem of forest fire detection using a wireless sensor network. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect