Abstract: With the rapid development of automation and artificial intelligence technology,the traditional manual inspection method of step-down substation and step-up substation substations will be replaced by robot inspection. Most of the existing substation inspection robots use a single navigation method,and these navigation methods have problems such as low precision and poor stability. An adaptive fusion method of vision and laser navigation is proposed. During the navigation process,the accuracy of the vision and laser navigation methods are judged respectively,and the judgement results are normalized. Fusion to obtain more accurate and stable robot pose information. After repeated experiments,using the mentioned method in the environment where the navigation line is covered,the positioning accuracy of the fusion navigation method proposed reaches 2. 0 cm;in the environment where the surrounding scene changes greatly and the laser navigation effect is poor,the positioning accuracy reaches 1. 7 cm,which can meet the positioning and navigation of intelligent inspection robots in complex environments.