Functional health prognosis approach for multistation manufacturing systems considering workpiece quality risk based on heterogeneous quality stochastic flow network

In multistation manufacturing systems, workpieces usually exhibit heterogeneous quality states caused by material inconsistency, machine degradation, process disturbance, and environmental fluctuation. These states could be transformed into workpiece quality risk through production processing, inspection, and usage stages, leading to rework, scrap, and infant failure. Unfortunately, existing studies often simplify workpiece quality risk as a single observable deviation or aggregated quality-loss index, while the effects of residual latent defects and coupling-induced defect evolution are insufficiently captured. To address this issue, this paper proposes a functional health prognosis approach for multistation manufacturing systems considering workpiece quality risk based on a Heterogeneous Quality Stochastic Flow Network (HQSFN). First, the formation mechanism of workpiece quality risk under workpiece quality heterogeneity is analyzed, and the functional health connotation is defined from the perspective of capacity–load matching under workpiece quality risk-induced feedback. Second, an HQSFN model integrating heterogeneous quality state transition, machine degradation, and task execution is constructed. Third, a recursive inference procedure is developed to estimate the functional health margin (FHM) and predict the remaining useful life (RUL). Finally, a case study of a nuclear reactor shielding assembly component manufacturing system is provided for verification.