The work aims to propose an improved efficient path search algorithm based on the A algorithm to solve the problems of low search efficiency, large corner and non-compliance with the operation constraints of campus unmanned distribution vehicles in the application scenario of campuses. Firstly, the unnecessary round-trip search was reduced by improving the specific calculation method of the evaluation function and adding a weight coefficient to the prediction function to improve the search speed. Then the vector angle and the line segment were calculated to eliminate the redundant nodes to reduce the total length of the path. The data storage structure of the OPEN table was replaced, and the minimum heap was used to replace the array to access the optimal node faster, so as to improve the computational efficiency of the algorithm. Finally, the Bessel curve was used to smooth the path, to smoothen the place with sharp corners and reduce the corners. The experimental results showed that the improved A algorithm improved the search efficiency by 58% compared with the traditional A algorithm, and also significantly improved the path length and path smoothness. The experimental results show that the improved A algorithm can be used for the path planning of campus unmanned vehicles, which verifies the feasibility of the algorithm.