The multi-degree-of-freedom robotic arm 3D printing can break through the limitation of the growth direction of a single model in the traditional plane layered fused deposition 3D printing process and support multi-directional printing.Surface layering and partitioning planning are the prerequisites for robotic arm printing.For the model with multi-pillar structure or large surface change rate,after surface layering,it will show surface structure features with multiple islands.The traditional partitioning method based on the association of space and plane mapping and the partitioning method based on surface curvature characteristics have problems such as loss of point and surface information,inability to accurately construct regional relationships,too many sub-regions and unreasonable regional division.Aiming at these problems,a complex multiple islands partitioning method based on surface layering was proposed.By traversing the information of each surface layer,the surface layer needed to be partitioned was judged;the breadth-first search algorithm was used to extract the point-surface information and contour information in the same closed area;according to the association of the point-surface information of the original surface layer,the newly stored patch information was updated and combined into a new surface to complete the area division.Finally,the two models with multi-pillar structure and large variation of outer surface were partitioned,and the feasibility of the partition method was proved by the simulation of path planning and the 3D printing experiment of the robotic arm.