An Improved RRT-Based Algorithm for Optimized Path Planning in Multi-Dimensional Environments

Objective: This study aims to improve the classical Rapidly-exploring Random Trees (RRT) algorithm, commonly used in robotic path planning. The classical algorithm often generates suboptimal paths due to its reliance on random exploration. Enhancements are proposed to increase pathfinding efficiency and applicability in multi-dimensional environments. Methods: An improved RRT algorithm was developed by introducing a target-biased tree expansion strategy and optimizing step sizes based on algorithmic principles and operational scales. Simulations were performed in two-dimensional (2D) and three-dimensional (3D) environments to evaluate its efficiency. Key performance metrics, including path length, computational time, and total number of explored points, were analyzed and compared with the traditional RRT. Simulations with dynamic obstacles were also performed in 2D environments in order to validate the performance of the improved algorithm in complicated and realistic situations. Results: The improved algorithm consistently generated shorter paths and reduced computational time compared to the classical RRT. For both 2D and 3D simulations, the success rate of finding optimized paths significantly increased, demonstrating enhanced adaptability in various environments. Conclusion: The proposed improvements address the limitations of the traditional RRT algorithm, providing a practical and effective solution for robotic path planning in static and multi-dimensional environments. These advancements contribute to the development of more efficient path-planning methods in robotics.