The assembly process cannot meet the increasing demands with the conventional method. To obtain an optimal assembly performance, the machined surface should be reconstructed, and further analysis should be conducted to provide a direct guidance for the assembling process. Surface topography can be represented by a sufficient number of points; however, the measurement of dense points would be time-consuming. In this study, the measured points are arranged in a nonuniform manner, and processed by interpolation for the reconstruction of the actual surface and the prediction of the assembly performance. Point clouds processed by 4 interpolation methods were used to verify the proposed method. The sampling strategy and data processing was concluded to be accurate; measurement time was significantly decreased compared to the dense points. It is proposed that the three indexes should be taken into consideration when the method is used to process the measurement data of different machined surfaces.

传统方法无法满足日益增长的组装需求。为了获得最佳的组装性能，应重建加工表面，并应进行进一步分析以为组装过程提供直接指导。表面形貌可以由足够多的点表示。但是，密集点的测量将很耗时。在这项研究中，测量点以不均匀的方式排列，并通过插值进行处理，以重建实际表面并预测装配性能。通过4种插值方法处理的点云被用来验证所提出的方法。结论是抽样策略和数据处理是准确的；与密集点相比，测量时间显着减少。