Abstract: Addressing the inability of existing automatic annotation methods for drawings to fulfill diverse practical requirements, and considering the similarity among steel structural components, a drawing cloning method is proposed. This method takes manually annotated drawings as the baseline and accomplishes automatic annotation for similar components through cloning. The proposed method consists of two parts: rule learning and annotation matching. The former is suitable for cloning annotations with general applicability, while the latter is tailored for non-general annotations. Implemented on the 3D3S Solid platform, this method was subsequently applied to the drafting work of a steel structure factory. By comparing it with the existing drawing cloning method on the Tekla Structure platform, the effectiveness of this method was validated. The annotation recall rate of the cloned drawings reached over 90%, slightly higher than Tekla Structure's 87.5%. Additionally, compared to Tekla Structure, this method can resolve the annotation cloning issue for components (or parts) with partial differences through repetitive cloning. Furthermore, it can reduce the manual drawing workload for repeated annotations in the same type of drawing through self-cloning.