Abstract: When the lunar rover in the future performs complex tasks such as shadowed regions detection and autonomous sampling and return under weak communication conditions in the lunar south pole, the demand for high-precision autonomous navigation system will become more urgent. Taking into account the safety guarantee of the lunar rover and the lunar environmental factors with navigation benefits, a new fusion path planning algorithm for high-precision autonomous navigation was proposed in this paper, by integrating a global auxiliary path based on optimized A* algorithm with Dynamic Window Approach. With location error of simultaneous localization and mapping based on lidar serving as an evaluation index, a numerical simulation and a semi physical experiment were established to verify the feasibility of this algorithm and the effectiveness of improving navigation accuracy of the rover in different application scenarios. Experiment results show that compared with traditional path planning algorithms with single factor considerations, the average absolute location error of simultaneous localization and mapping algorithm in the two-dimensional plane was reduced by a maximum of 42% when the lunar rover moved along the path planned by the proposed algorithm, which can provide technical support for autonomous navigation and path planning tasks in the complex environment of the lunar south pole.