曲面加筋结构拓扑优化密度场结果，存在不连续、过小结构特征等问题，难以直接应用于后续精细设计及加工制造，而基于人工经验进行的特征提取及模型重构，存在操作繁琐、重构周期长等问题。针对上述问题，提出了一种面向曲面加筋结构拓扑优化结果的参数化重构方法。首先，针对曲面加筋拓扑优化结果，建立一种三维曲面空间与二维平面空间的映射关系，并基于正向网格映射实现曲面优化结果向平面优化结果的转换；然后，针对平面优化结果，基于图像形态学方法实现轮廓特征参数化提取，并采用样条曲线插值方式实现几何模型重构，获得平面优化结果参数化模型；最后，基于逆向映射网格映射方法，获得曲面优化结果参数化模型，实现曲面加筋结构拓扑优化结果的参数化重构。基于所提出的方法，以承力筒、舱门及密封舱这3种典型曲面加筋结构为对象开展算例研究，并将重构结果与优化结果进行对比。结果表明，参数化重构模型的结构响应与拓扑优化结果误差在5%以内，表明提出的方法具有优异的重构精度。

The topology optimization density field results of curved stiffened structures have problems such as discontinuity and too small structural features， which make them difficult to be directly applied to subsequent fine design and manufacturing. Meanwhile， feature extraction and model reconstruction based on artificial experience have problems such as cumbersome operation and long reconstruction period. To address the above problems， a parametric reconstruction for topology features of curved stiffened structures method is proposed. Firstly， a mapping relationship between three-dimensional surface space and two-dimensional planar space is established for the results of surface stiffened topology optimization， and the transformation from surface optimization results to planar optimization results is realized based on forward mesh mapping. Then， for the planar optimization results， the contour feature parameters are extracted based on the image morphology method， and the geometric model is reconstructed by spline curve interpolation to obtain the parametric model of the planar optimization results. Finally， based on the inverse mapping mesh mapping method， the parametric model of the surface optimization results is obtained， and the parametric reconstruction of the topology optimization results of the surface stiffened structure is realized. Based on the proposed method， three typical curved stiffened structures of bearing cylinder， cabin door and sealed cabin are taken as examples， and the reconstruction results are compared with the optimization results. The results show that the error between the structural response of the parametric reconstruction model and the topology optimization result is within 5%， indicating that the proposed method has excellent reconstruction accuracy.