The Theta* algorithm utilizes an eight-neighborhood heuristic for path search, generating numerous redundant expansion nodes that impair search efficiency. Meanwhile, the traditional artificial potential field (APF) algorithm suffers from local optima and target unreachability issues. To address these limitations, this paper proposes a hybrid path planning framework integrating three core components—JPS-Theta* algorithm, B-spline smoothing, and improved APF algorithm—with clear synergistic interactions. Specifically, the JPS-Theta* algorithm fuses jump point search (JPS) with Theta* to eliminate redundant nodes and generate a globally optimal path: JPS prunes non-essential intermediate nodes to boost search efficiency, while Theta*’s line-of-sight check ensures path smoothness. Next, B-spline curves are applied to further smooth the JPS-Theta* path, eliminating sharp corners to meet the robot’s kinematic constraints. Finally, the improved APF algorithm incorporates global path control nodes (extracted from the smoothed B-spline path) to modify attractive and repulsive potential functions, enabling real-time dynamic obstacle avoidance while maintaining alignment with the global optimal trajectory. Simulation results using MATLAB and ROS show that the two algorithms effectively integrate. Compared with the original Theta* algorithm, the JPS-Theta* algorithm shortens the robot’s path search time and reduces a substantial number of expansion nodes, thus realizing comprehensive excellent performance, including fast path search, fewer redundant nodes, short path distance, and smooth path. Meanwhile, in comparison with the traditional APF algorithm, the improved APF algorithm generates a shorter local path with less time consumption, and the local path exhibits a closer fit with the global path.
Li et al. (Tue,) studied this question.