Leveraging RRT*: Probabilistically Interpreted Mechanisms Enhanced With P-HOPE and FLEX-OPT for Complex Path Planning

Path planning is a great challenge in the autonomous navigation of mobile robots. The Rapidly-exploring Random Tree $^{\ast }$ (RRT $^{\ast } $ ) algorithm is widely used for its probabilistic completeness. In the literature, improved RRT $^{\ast }$ -based algorithms usually enhance search efficiency through different target bias strategies. However, these algorithms often fall into obstacle traps in complex environments with narrow passages or high obstacle densities due to the loLcal minima problem in the optimization process. In addition, the existing algorithms also exhibit inefficient sampling and slow convergence in large-scale maps. To tackle with these problems, we propose an improved algorithm, called the RRT $^{\ast }$ -PRIME (Probabilistically Interpreted Mechanisms Enhanced RRT $^{\ast } $ ) algorithm, in this paper. First, a powerful strategy, called the P-HOPE (Probability-Driven Heuristic Optimization for Path Exploration) strategy, that integrates multidimensional influencing factors is designed in the proposed RRT $^{\ast }$ -PRIME algorithm to optimize target sampling direction by considering angle, direction consistency, and obstacle distribution. Second, a flexible mechanism FLEX-OPT is developed to adaptively and dynamically adjust the search strategy through real-time feedback and monitoring of the cost function to tackle the above-mentioned local minima problem, which significantly improves the convergence speed and path quality of the algorithm. The experimental results suggest that the proposed RRT $^{\ast }$ -PRIME algorithm can reduce the initial solution search time by 76.32%, reduce the number of search nodes by about 80.67%, and improve the search path quality compared with the RRT $^{\ast }$ algorithm. In both narrow complex and large-scale map environments, the RRT $^{\ast }$ -PRIME algorithm significantly outperforms the RRT $^{\ast }$ , Informed-RRT $^{\ast }$ , h-RRT $^{\ast }$ , and PF-RRT $^{\ast }$ algorithms in terms of reliability and efficiency. These results demonstrate the effectiveness of the RRT $^{\ast }$ -PRIME algorithm as a robust and efficient solution for path planning in complex and large-scale environments.