Aiming at the problem that the wall climbing robot is difficult to realize complete coverage detection on the outer wall of the oil tank, a priority heuristic path planning algorithm based on rolling window was proposed. The grid map was used to model the environment in the rolling visual window, and the priority heuristic algorithm was used to search coverage path in rolling planning window. For the existence of U-shaped obstacles in the environment, the recognition and direct filling of U-shaped obstacle were carried out. When the robot entered the dead zone, the dead zone escape algorithm was enabled to make the robot escape the dead zone smoothly. The simulation results showed that the path planning method could guide the wall climbing robot to achieve complete coverage efficiently under the condition that the outer wall environment of oil tank was unknown. The research results have certain theoretical and engineering significance for improving the coverage efficiency of the wall climbing robot on the outer wall of oil tank.