To address the issues of a large number of search nodes and slow planning speed in the traditional Hybrid A* algorithm, an improved Hybrid A* algorithm is proposed that adaptively adjusts the node expansion strategy based on obstacle distribution. Within a predefined annular search area, the algorithm evaluates the presence of obstacles. If no obstacles are detected, it introduces the angular deviation between the current node and the goal, executing path planning tasks using a goal-oriented strategy. Conversely, upon detecting obstacles, the algorithm dynamically calculates the node expansion step size according to the obstacle distribution structure, replacing the fixed step size in the original algorithm. Additionally, to enhance the smoothness of the path, a consistency judgment function among nodes is introduced to identify and remove similar nodes. Experimental results show that the improved Hybrid A* algorithm not only adapts to various complex obstacle terrains but also, compared to the original algor