DBVSB-P-RRT*: A path planning algorithm for mobile robot with high environmental adaptability and ultra-high speed planning

To address the limitations of the rapidly-exploring random tree star (RRT*) algorithm, such as slow convergence, high time cost, and weak environmental adaptability, which have hindered its application in the field of mobile robot path planning, this paper introduces a bi-directional P-RRT* algorithm with adaptive direction biased and variable-step-size (DBVSB-P-RRT*). In the sampling phase, anew adaptive target deviation sampling mechanism is designed to alter the sampling angle during the planning process based on the obstacle configuration between the continuously iterated new node and the target point. In the RRT* extension phase, improvements are made to the attractive potential field (APF) algorithm, and an adaptive step size is devised to enhance environmental adaptability and guidance by integrating with the RRT* algorithm. Additionally, a collision detection function has been designed to preemptively eliminate a portion of obstacles, thereby reducing the number of iterations and enhancing the efficiency of the planning process. In the RRT* connection phase, a bidirectional search strategy is introduced to boost the planning speed of the RRT* algorithm. Simultaneously, path pruning is implemented to eliminate redundant nodes and optimize path quality. Finally, comparisons are made with the RRT*, Bi-RRT*, Bias-RRT*, Bias-P-RRT*, and Informed-RRT* algorithms in seven different environments, including comparative data and analysis, validating the feasibility and superiority of the new optimization algorithm. Through comparison, the new algorithm demonstrates significant improvements, it yields shorter and smoother path lengths compared to other algorithms. In seven environments, it maintains a 100% success rate, reducing the required time by at least 40%. Overall, it exhibits superior path quality, enhanced adaptability to varying environments, and ultra-high speed.