A key process identification framework for aircraft assembly production based on the network with physical attributes

Accurate identification of aircraft assembly key processes plays an important role in aircraft production management. However, due to complex processes, multiple attributes, and the aggregation phenomenon of the aircraft assembly process, identifying the key processes faces huge challenges. Therefore, a network-based key process identification framework is proposed in this paper. Firstly, according to assembly processes and vital physical attributes, an aircraft assembly network and the node attribute matrix are constructed. Then, the SC-Qwalktrap algorithm is designed to adaptively identify the aircraft assembly network community structure. Subsequently, the network-based influential node identification algorithm is proposed to recognize key process nodes, which consists of two steps. Within the community, local influence is evaluated based on node entropy and network topology. Between the communities, global influence is measured based on neighboring nodes in different communities. Finally, the proposed framework is compared with the traditional centrality measurements on the datasets from PSPLIB and commercial aircraft assembly datasets. The experiment results demonstrate that the network-based influential process identification algorithm can effectively identify the key processes.