Termografía Infrarroja aplicada a la detección de incendios en la interfaz urbano-forestal y su optimización mediante redes neuronales artificiales
[EN] The Albufera of Valencia and its Devesa, form a single unit with both an ecological and social high value; reason that led them to be declared Natural Park in 1986 by the Generalitat Valenciana; being the first park declared in this autonomous community. The Devesa is the spit that separates th...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Fecha de publicación: | 2015 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | español |
| OAI Identifier: | oai:riunet.upv.es:10251/49830 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/49830 |
| Access Level: | acceso abierto |
| Palabra clave: | Incendio Forestal Sistema Detección Incendio Termografía Infrarroja Interfaz Urbano Forestal Redes Neuronales ANN Índice de Riesgo Clasificación Alarmas Devesa Albufera INGENIERIA HIDRAULICA |
| Sumario: | [EN] The Albufera of Valencia and its Devesa, form a single unit with both an ecological and social high value; reason that led them to be declared Natural Park in 1986 by the Generalitat Valenciana; being the first park declared in this autonomous community. The Devesa is the spit that separates the Mediterranean Sea from the Albufera lake, and is considered a natural area with high scientific, cultural, scenic and educational values. And although during the 60s underwent a process of urbanization, today is in the throes of regeneration to an era of ecological climax. This regeneration has been possible because of the economics efforts made by various administrations for their conservation and protection. However, these efforts do not prevent Devesa from suffering systematically wildfires that undermine their ability to regenerate and that not only produce significant ecological and economic damage, but when fires reach great dimensions, threaten the life of the people who live there and also the firefighting services This Thesis focuses on the study and optimization of the detection system of wildfires using infrared installed in the Devesa. For doing this, the wildfires produced during ten years and the alarms generated during five years of operation of the system are analyzed, relating these alarms with the weather conditions; in order to reduce false positives; on the other hand a fire risk classification system based on neural networks is developed, using as descriptions parameters those used in the IFW, that is the official system used for the official organization in charge of fire weather index in Spain: AEMET. After the development of the neural network to classify the risk of fire, and analyzed the infrared camera system, both are combined to establish a classification system of the alarms, in order to reduce false positives, and establish a criterion of risk to the user of the fire detection system. |
|---|