To address the issues of excessive search nodes, complex paths, and low efficiency in traditional A* algorithms for UAV path planning, this paper proposes a novel A* UAV path planning algorithm based on an optimized heuristic function and B-spline curves. By constructing a new heuristic function incorporating obstacle loga-rithmic factors and dynamic weight coefficients, the algorithm effectively reduces the number of search nodes in global path planning. To balance UAV dynamic performance and computational efficiency, the traditional 8 search directions are streamlined to 5, thereby lowering computational complexity. An improved Floyd algorithm is employed to eliminate redundant intermediate nodes in the path, minimizing unnecessary turns. Additionally, a cubic B-spline interpolation algorithm is applied to smooth the planned path, ensuring smoother flight tra-jectories. Comparative experiments under two different map environments demonstrate that the improved A* algorithm outperforms the traditional A* and ant colony algorithms, generating more efficient and smoother paths.