Analysis and Contributions of Classical Techniques for Path Planning

• Lidia Rocha UFSCar
• Kelen Vivaldini UFSCar

Resumo

Unmanned aerial vehicles have been increasingly used to solve problems in several areas. These problems require the UAV to move between two nodes precisely to perform the tasks. In this way, path planning algorithms are adopted to provide this path. The classic algorithms can confirm whether a result exists or not and be great for flying in real-time. However, there are many classical algorithms, and each of them has better performance in different environments. Therefore, this paper presents a comparison and statistical analysis between the classical algorithms, emphasizing the best algorithms, showing the advantages and disadvantages of each algorithm in indoor and outdoor environments. The algorithms that presented better performance are tested in a 3D environment. The analyzed algorithms are A*, Artificial Potential Field, Probabilistic Random Map, Rapid Random Tree, and Rapid Random Tree Connect. We validate the results in 2D and 3D simulations, in Python, and on Gazebo.

Palavras-chave: Solid modeling, Three-dimensional displays, Statistical analysis, Probabilistic logic, Unmanned aerial vehicles, Robustness, Real-time systems