Abstract: In view of the advantages of the indoor air disinfection robot, which has the advantages of liberating the labor force, reducing the risk of infection, and making the disinfection work flexible and intelligent, a set of BIM-based air disinfection algorithm is designed, the robot disinfection work mode is determined, and the information of menbers in IFC is extracted. Reconstruct the indoor model, improve the boustrophedon decomposition method to achieve regional decomposition, design the best reciprocating direction of the region, and determine the connection path between regions based on the visual graph and immune optimization algorithm. The simulation experiment proves the feasibility of the model and the algorithm. The comparison between the design and the A* algorithm shows that the visual graph algorithm has more advantages in the transfer distance and transfer time.