With the rapid development of network information technology,it has promoted the development of manufacturing revolution characterized by manufacturing informatization and intelligence.The research of advanced manufacturing mode,which integrates traditional manufacturing mode and network information technology deeply,has become the focus of research by experts and scholars in academia and industry.By learning the concepts of cloud computing in the manufacturing field,cloud manufacturing emerged as a service-oriented networked manufacturing model.The research on service composition optimization in the cloud manufacturing environment has achieved important practical value in integrating social manufacturing resources and improving resource utilization efficiency.At the same time,it has also generated the problem of matching resource services and user needs: how to match users with satisfactory resources in response to large-scale demand in real life.Therefore,in view of the limitations of traditional matching mode optimization capabilities under large-scale demand,the service composition optimization problem of cloud manufacturing is studied from two levels of technology improvement and case application to promote the effective allocation of resources.The main research results and innovations of the thesis are as follows:(1)Proposing an evaluation index system for large-scale demand in the cloud manufacturing environment.First,the process of cloud manufacturing resource optimization under large-scale demand is proposed,and the key process from the perspective of large-scale demand research of this thesis is clarified.Second,the two types of evaluation indicators are determined.One is the general evaluation of Qo S(Quality of Service,Qo S)proposed in combination with the existing literature.The second is to propose two unique indicators based on the research problem,and carry out a reasonable indicator measurement,which effectively measures the constraints generated by the resource service composition process under large-scale demand.Finally,according to the modeling ideas of this article,determine the weight calculation method in the transformation process.(2)Constructing a multi-objective optimization model without supply constraints under large-scale demand.Based on the analysis of the limitations of traditional modeling,relax the restrictions on supply constraints and allow more resources and services to participate in the optimal selection.At the same time,based on the background of largescale demand problems,conventional constraints and specific constraints are determined,and a multi-objective optimization model is established.It breaks through the limitation that the supply is greater than the demand in the traditional service composition optimization model,and provides more solution options for the research of service composition problems under large-scale demand.(3)Designing an adaptive differential evolution algorithm(SMADE)based on stable mutation.In view of the uncertain variable process of differential evolution algorithm and the limitation of invalid mutation,a mutation strategy based on DE/best-worse/1 is proposed;In view of the limitation of poor adaptive mutation rate of differential evolution algorithm,a crossover rate based on growth rate is proposed,which improves its flexibility.In order to verify the effectiveness of the SMADE algorithm,a test function is used to compare the SMADE algorithm with several classic differential evolution algorithms.The experimental results show that the SMADE algorithm has better results in convergence speed and solution accuracy.(4)Analyzing the influence of model parameters on the solution of the optimization model,and explored the potential laws of the service composition optimization model.According to the characteristics of the problem,a two-stage solution strategy based on the SMADE algorithm is proposed and applied to case analysis and numerical analysis.On the one hand,take the large-scale task requirements of a certain type of manufacturing enterprise as an example to verify the effectiveness of the proposed optimization model,and from the results of the examples: the fitness value of the optimization model proposed in this thesis is better than the results under traditional modeling;On the other hand,numerical experiments are used to reveal the influence of evaluation indicators on the optimization model,and to further study the potential laws of the optimization model.The experimental results show that the optimal solution of the model can be obtained when the index parameters are controlled within a reasonable range;In addition,compared with traditional modeling,the optimization model for large-scale demand is more extensive and general. |