Candidate service composition.

To address the issue of insufficiently comprehensive representation of service composition indexes in the shared manufacturing environment, service reliability, confidence, and other indexes are decomposed in detail to establish a composition evaluation system, and a shared manufacturing service composition optimization model based on bi-level programming is proposed. The model takes Quality of Service (QoS) as the upper objective function and service reliability, service confidence and task fit as the lower objective function. The upper objective function needs to be minimized, while the lower objective function needs to be maximized. To achieve the optimal composition scheme of shared manufacturing services, the Criteria Importance Though Intercrieria Correlation (CRITIC) is used to determine the weights of the indicators, and the improved Fast Elitist Non-Dominated Sorting Genetic Algorithm (Improved NSGA-II) is employed to solve the multi-objective optimization problem. Finally, the improved NSGA-II has a 23.33% increase in convergence speed and a 69.99% gain in operational efficiency when compared to the traditional NSGA-II. The viability and effectiveness of the improved NSGA-II have been demonstrated.