A Method for Identifying Error Sources in Complex Assembly Process Based on Multi-Level Network

In complex assembly process, there are many error sources which influence the final product quality, so the identification and compensation of error sources are the important ways to improve product quality. The purpose of this paper is to identify the error sources by establishing the relationships between assembly errors and product performance. Firstly, a new concept of multi-level network is proposed to describe the complexity of product structure. Then the multi-level assembly network (MLAN) model is established and its self-organization stability is discussed to explore the influences from assembly errors on product performance. The sensitive fluctuation index (SFI) is defined and calculated to identify the error sources in assembly process. Based on this, the error sensitivity is discussed so that measures for error compensation are made to improve the assembly quality of product. Finally, an assembly experiment about blade rotation is conducted to verify the feasibility of proposed method.