基于空间探索引导双树RRT*的路径规划OA

To tackle challenges such as sampling blindness,limited guidance strategies,and kinematic constraints inherent in the sampling-based RRT* path planning algorithm,this paper introduces an enhanced Space Exploration Guided Bidirectional RRT*(SGB-RRT*)path planning algorithm.The proposed method innovates by implementing a spatial exploration sampling strategy during the initial sampling phase.This approach actively investigates the environment surrounding the current node to establish comprehensive obstacle perspectives,thereby mitigating occlusion issues.This diminishes the blindness associated with sampling.In the subsequent guidance phase,the SGB-RRT* adopts a decay exploration mechanism coupled with a weight control strategy.The decay exploration dynamically adjusts its parameters based on the cumulative number of samples,progressively narrowing the exploration scope to concentrate efforts on guiding the sampling process toward the target region.Concurrently,distance weight is used to regulate the tendency degree.During the extension phase,the Dubins curves with the CCC type excluded are integrated with the bidirectional tree structure to address the kinematic constraints.In lieu of the conventional Rnear-based backtracking,it employs parent node backtracking.As a result,the algorithm efficiently generates low-cost,feasible paths that respect the kinematic constraints.Simulation experiments conducted across diverse maps compare the performance of SGB-RRT* against RRT,RRT*,and RRT*-connect algorithms.These tests confirm the superior computational efficiency,improved path quality feasibility of the proposed SGB-RRT*approach.