Service Composition Modeling and Optimization Method for QoS Evaluation of Intelligent Manufacturing Supply Chain under Ripple Effect

Aiming at the problem of QoS（Quality of Service）evaluation inaccuracy and ripple effect transmission delay caused by insufficient dynamic regulation of service time attributes in intelligent manufacturing supply chain, this study proposes a resilient Manufacturing Service Composition And Optimization Selection method (IMSCOS) based on QoS time attribute lean evaluation. Firstly, the limitations of the preset parameter fixation and linear response mechanism of the traditional resilience management model are broken through. By deconstructing three-dimensional indicators such as service cycle elasticity coefficient, time window dynamic fault tolerance threshold and task coupling delay sensitivity, a service composition resilience evaluation model with spatiotemporal correlation characteristics is constructed. On this basis, a multi-objective optimization algorithm integrating time constraint relaxation mechanism and chaotic adaptive perturbation strategy is designed to achieve a dynamic balance between Pareto front convergence efficiency and solution space distribution diversity. Experimental results show that compared with traditional optimization methods, the IMSCOS method improves key indicators such as manufacturing cycle volatility and service reliability index by 28.6% and 19.4% respectively, and shortens the service composition reconstruction response time by 32.8%, effectively improving the resilience buffering capacity of the supply chain system to time attribute disturbances. This study provides a solution with both theoretical value and engineering application prospects for building a spatiotemporal collaborative service composition decision system for intelligent manufacturing supply chain.