Abstract: Complex products are multi-process manufacturing parts (MMP),such as aero-engine structural parts and automobile bodies.The product quality control in a dark state is caused by the coupling impact of the generation and accumulation of the various errors in the manufacturing process on product quality.Constructing the models of the stream of vairiation (Sov) to describe and analyze the transmission,accumulation,and elimination of variations through combining the theory of Sov with precision control theory and engineering data processing methods is important to achieve quality control for the production of complex products in MMP.This paper summarizes the Sov models application and other supplementary researches,puts forward the overall framework of quality control of SOV models for MMP.It is classified from two aspects: small data-driven time series mathematical analysis models and big data-driven complex network statistical analysis models.Finally,the development trend of Sov models in the field of MMP of high-end equipment and complex products in the future is prospected.

Key words: complex product, multi-process manufacturing, quality control, stream of variation

摘要： 航空发动机结构件、汽车车身等复杂产品通常为多工序制造件,其制造过程中多种误差的产生和累积会对产品质量产生耦合影响,使产品质量控制处于暗箱状态。将误差流理论与精度控制理论、工程数据处理方法相结合,构建对误差的传递、累积、湮灭进行描述和分析的误差流模型对于多工序制造复杂产品质量控制具有重要意义。本文总结和归纳了误差流模型的构建方法及其应用研究,提出了多工序制造的误差流模型质量控制总体框架,从小数据驱动的时间序列数学解析模型和大数据驱动的复杂网络统计分析模型两个方面对多工序制造的误差流模型进行分类综述。最后,展望了误差流模型在高端装备和复杂产品等未来多工序制造领域中的发展趋势。

关键词: 复杂产品, 多工序制造, 质量控制, 误差流