Karaelmas International Science and Engineering Symposium (KISES 2024), Zonguldak, Turkey, 10 - 11 May 2024, pp.67
Nowadays, the use of robots has become widespread due to the many benefits it provides. Robots are expected to avoid static or dynamic obstacles and move by kinematic and dynamic constraints while performing the given task. One of the most important problems to be solved in robots is the path planning problem. There are many algorithms and methods used to solve this problem. The choice of these algorithms and methods depends on factors such as the physical structure of the robot and whether the robot performs its task in an indoor or outdoor environment. In this study, the Rapidly exploring Random Tree (RRT) algorithm was used to solve the optimum path planning problem of robots in a two-dimensional environment. The RRT algorithm is very successful in quickly creating a path from the starting point to the target point by creating a probabilistic tree structure. The algorithm randomly determines a new node in each iteration, and the new node to be added to the path is determined according to the closest point on the tree to the determined new node. Due to the concept of probabilistic completion, the RRT algorithm guarantees that a possible solution will be reached if sufficient resources are available. However, it does not guarantee an optimum running time in reaching the solution. The RRT algorithm can move probabilistically in all directions and points in the configuration space. The RRT algorithm is also asymptotic optimal, giving better results as the difficulty and size of the problem increase. Asymptotic optimality may not always give the best solution, but in general, it shows that the performance of the algorithm tends to improve. With these features, the RRT algorithm is preferred in complex and dynamic environments. The RRT algorithm used in the study was run separately on CPU and GPU architectures, and the results were compared in terms of runtime.