Um método de planejamento de rotas de voo de VANTs multirotor para cobertura de áreas utilizando a meta-heurística ACO

• Ano de defesa: 2019
• Autor(a) principal: Franco, Lucas dos Santos
• Orientador(a): Kato, Edilson Reis Rodrigues
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: Agricultura de precisão; Planejamento de rota de cobertura; CPP; Problema do caixeiro viajante generalizado; GTSP; Otimização por colônias de formigas; ACO; UAV
• Palavras-chave em Inglês: VANT; Precision agriculture; Coverage path planning; Generalized travelling salesman problem; Ant colony optimization
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::MATEMATICA DA COMPUTACAO; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::METODOLOGIA E TECNICAS DA COMPUTACAO; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/12177

With the popularization of Unmanned Aerial Vehicles (UAVs), and with the expansion of the application areas of this technology, there is an increase of research focused on flight route planning. This work starts from the use of UAVs in agricultural scenarios in the aerial imaging task. The objective is to present a Coverage Path Planning (CPP) method for multirotor UAVs and compare it with a solution already used in the market. The developed method considers scenarios with multiple terrains, and seeks to present a route proposal with optimization of the order of visitation of the terrain, minimizing the number of curves of the route and the total distance of the route. The presented method contains three main steps. The first deals with the decomposition of areas, where terrains represented by concave polygons are decomposed into smaller convex shaped subareas using a greedy algorithm. The second step calculates the flight direction that minimizes the number of course curves in each of the subareas, finding the direction of the highest polygon height to guide the direction of the round-trip movement pattern application, known in the literature by default boustrophedon. The third step deals with subarea visit order optimization, at this stage the scenario is modeled as a specialization of the Generated Travelling Salesman Problem (GTSP), and to solve this problem we use the Ant Colony Optimization algorithm (ACO). The results obtained by the proposed method are compared with the solutions proposed by a route planning program already used by the market, the Mission Planner. To measure the efficiency of the solutions, two variables were considered: the total distance traveled and the number of route curves. Through the obtained results one can map the types of scenarios where the developed method can aggregate with the market.