Hypergraph-based soft community detection enabling interpretable reconfiguration in flexible assembly systems
提出一种基于超图的软社区检测方法，通过建模资源间高阶协作关系并融合语义属性，实现柔性装配系统中可解释的资源划分与瓶颈动态分析，为管理者提供需求驱动的配置策略。
The reconfiguration of resources in flexible assembly systems (FAS) plays a crucial role in accelerating new product introduction. While existing resource allocation methods primarily aim to enhance system performance, they often lack interpretability regarding the underlying configuration patterns. To address this gap, this paper introduces a hypergraph-based soft community detection approach for resource configuration in FAS, which explicitly accounts for the dynamic evolution of resource bottlenecks under different reconfiguration schemes. The proposed method models high-order collaborative relationships among resources via a hypergraph and leverages meta-paths to extract semantic collaboration patterns, thereby enriching the information used in soft community detection. A hybrid probabilistic generative model is developed to integrate structural collaboration logic with semantic attributes, enabling interpretable and adaptable resource partitioning. Bottlenecks are automatically identified through a joint analysis of structural centrality and resource flexibility. Experiments on synthetic datasets and a real fuselage wall-panel assembly system demonstrate the effectiveness and advantages of the proposed framework. This approach offers managers a demand-driven resource configuration strategy while providing actionable insights into bottleneck dynamics.