Path planning is a critical task in automated navigation and seeks to identify optimal collision-free routes for autonomous systems such as unmanned vehicles, aircrafts, and surface ships within a specified environment. The rapid computation of optimal paths in large-scale grid maps remains a major challenge. This paper presents an enhanced HPA pathfinding algorithm that utilizes an abstract representation of grid maps to facilitate fast navigation. The enhancements to the HPA algorithm include a detailed examination of neighborhood branch extensions, the incorporation of a high-quality heuristic function, and the implementation of repulsive force fields. The effectiveness of this improved HPA algorithm in large-scale grid environments is demonstrated by extensive experiments in real-world settings. In terms of computational efficiency, for path planning in real scenarios, the computation time of the HPA Improved algorithm can be reduced by more than 95% relative to the A algorithm, and the computation time can still be reduced by more than 80% relative to the HPATheta algorithm, even though the computation time can be reduced by more than 80%. Even compared to HPATheta algorithm, the computation time can still be reduced by more than 80%. Specifically, the improved HPA algorithm significantly reduces the time required to generate path results while also enhancing path safety. This study provides valuable insights into advanced pathfinding techniques that could advance automated navigation systems.