Tensor algebra-based parameter correlation modelling and analysis for variant design of open architecture production line towards resilience

The variant design of open architecture production lines has emerged as a robust and agile approach to swiftly adapt to demand fluctuations or disruptions, leveraging its inherent advantages of cost-effectiveness and the rapid redeployment of best practice knowledge. Within this context, parameter correlation modelling of the open architecture production line is an indispensable step in the variant design process. This paper introduces a novel method for parameter correlation modelling within a multi-layer heterogeneous network of production lines, utilising tensor algebra and hyper-adjacency matrices. This technique meticulously analyzes and uncovers the latent information among parameter nodes, thereby significantly enhancing the reuse of collaborative design knowledge. Moreover, it establishes topological indicators for parameter nodes in the multi-layer heterogeneous network, enabling the calculation of the significance of each parameter node. This is instrumental in analyzing the potential disruption impact and the disruption propagation path of different parameter nodes. To substantiate the viability of the proposed method in bolstering system resilience, a case study of a smartphone assembly line is presented. This case constructs a parameter correlation model of the smartphone assembly line and applies node significance calculations to derive the potential change-propagation paths when different nodes are disrupted.