In industrial internet environments, industrial software systems are usually large-scale and with a complex and heterogeneous array of software systems and physical devices. Based on the unified service-oriented and component-based encapsulation of heterogeneous devices and software objects, diverse applications of industrial software systems can be flexibly constructed through the orchestration of service components. However, it usually requires high efficiency of service orchestration and the assurance of service quality. To enhance the efficiency of service component orchestration and ensure service quality in industrial internet environments, this paper proposes a service component orchestration method called FGraphSAGE_GA, which formalizes the convert the component orchestration problem in industrial internet environments into a link prediction problem within a graph structure. It employs a supervised learning approach to train the prediction model. For the new service component orchestration problem, the trained model performs probabilistic predictions to narrow down the candidate space, followed by the optimization of the service component orchestration plan using a genetic algorithm. Experiments were conducted on two different datasets for service component orchestration problems of three different scales, and experimental results show that the proposed FGraphSAGE_GA algorithm has good performance in terms of service quality and orchestration efficiency.

The existing expensive constrained multi-objective optimization algorithms based on surrogate models face two main issues. Firstly, the use of regression models to fit constraints introduces errors that affect the algorithm's search direction. Secondly, when the objective function is non-fittable, the performance of the regression model for fitting is poor. To address these issues, a collaborative expensive constrained multi-objective evolutionary optimization algorithm is proposed, which combines a classification model with a regression model. This method employs the classification model to roughly divide the search space, guiding the algorithm to quickly enter the feasible region and reducing the impact of constraint fitting errors. The regression model is then used to optimize the objective function within the feasible region. The collaboration of the two models allows the classification model to provide a general search direction while the regression model performs detailed modeling. This fusion of models not only considers the impact of constraint errors on the algorithm but also comprehensively addresses the fittability of the objective function, enabling a more comprehensive and accurate depiction of the characteristics of complex problems. As a result, it enhances the efficiency and effectiveness of the algorithm, providing an effective approach for further improving expensive constrained multi-objective optimization based on surrogate models.