A Hybrid Path Planning Method Based on Improved A* and CSA-APF Algorithms

In order to address the challenges of global path planning in complex and dynamic environments, where avoiding dynamic obstacles is difficult, and local paths are prone to getting stuck in local optima, this paper proposes an enhanced hybrid path planning approach based on improvements to the
$\text{A}^{\ast} $
algorithm and the Artificial Potential Field (APF) algorithm. Firstly, addressing the issues of redundant path nodes and non-smooth paths generated by the
$\text{A}^{\ast} $
algorithm, we introduce enhancements such as weighted heuristic function, removal of redundant path nodes, and the incorporation of a cubic quasi-uniform B-spline curve to improve search speed and generate smoother paths. Secondly, addressing the problems of unreachable target points and susceptibility to local optima associated with the APF algorithm, two key strategies are employed. On one hand, the repulsive potential field function is modified by incorporating the m-th power of the relative distance from the robot to the target point to ensure reachability. On the other hand, the CSA-APF (Constrained Simulated Annealing with Augmented Potential Field) algorithm is introduced. This algorithm combines the APF algorithm with Simulated Annealing (SA), incorporating constraints on angles and safety distances to escape local optima and generate globally optimal points. Finally, the proposed methods are compared through simulation experiments in different scenarios. The experimental results demonstrate that the proposed approach effectively achieves path planning in complex dynamic environments.