Simulação de modelo de combate a incêndios florestais utilizando enxame de VANTs

Abstract

Approximately 31% of the planet’s surface is covered by forest area, serving as an important resource for a quarter of the global population and the economy. One of the main threats to this resource are wildfires. It is estimated that even after 15 years from burning, forests show no signs of recovery. This work proposes a model for combating forest fires as a complement to the existing methods of combat. The model involves the use of a swarm of drones and a control base. A group of surveillance drones traverses an area in search of detecting any signs of a fire. If detection occurs, a signal with fire data is sent via a LoRaWAN network to the control base. The base then sends suppression drones to the fire location. The work focuses on the suppression task, estimating, through simulations, the impact of drones on fires. Additionally, a LoRaWAN network was also set up using single-board computers as nodes in the model. To achieve these objectives, the following steps were taken: research with firefighters and analysis of the responses obtained, analysis of the water flow necessary for the complete fire extinction, simulation of suppression, development of the LoRaWAN network, and analysis of the results obtained. The research with firefighters revealed that they prefer the use of drones for passive tasks (monitoring and cargo transport). The results obtained in the analyses and simulations demonstrate proximity to what would likely occur in real scenarios and the possibility of using swarms as an auxiliary tool to other combat methods, especially in the early stages of a fire. Also, a possible architecture for the LoRaWAN network was shown. As future work, it is intended to add new variables in simulations, development of the task of detecting fire foci, and new tests on the LoRaWAN network.