Implementação de algoritmo de planejamento de caminho de cobertura com regiões heterogêneas para veículos aéreos não tripulados

Abstract

The lower cost and popularization of technologies such as Unmanned Aerial Vehicles enable the spread of Precision Agriculture, as it is more accessible than other forms of data collection and which can fly partially or completely autonomously. This is possible due to coverage path planning algorithms, which trace vehicle routes to a specific area. However, available algorithms have limitations, not allowing regions with different flight characteristics and collecting data with other properties. This work aims to implement a coverage path planning solution that considers regions with heterogeneous flights to collect images at different altitudes during the same mission. Changing the flight altitude closer to the ground makes it possible to collect a greater amount of data from the surface being observed. However, a flight entirely at a low altitude adds resources that are used during the mission. Therefore, a heterogeneous flight with multiple altitudes is a way to optimize the quality of the data collected by reducing unnecessary resource consumption. The development uses the DJI’s multirotor aircraft and its development tools, allowing the implementation of a controller application, with GPS information, that can plan and send coordinates referring to the path that the vehicle must travel. To validate the method, coordinates of agricultural crops with different shapes and dimensions were used, demonstrating the results in different locations for traditional homogeneous coverage flights and heterogeneous overflights with one or multiple special subareas at low altitude. By evaluating these scenarios, it was found that heterogeneous coverage path planning can result in an optimized path with a length up to 35.4% shorter than a homogeneous low-altitude flight, with the entire process taking a few milliseconds to execute.