Research On Cloud Manufacturing Service Match Optimization And System Implementation

Combining key techniques of manufacture with advanced technologiessuch as cloud computing and Internet of Things(IOT),Cloud manufacturing is a new kind of “Internet+ manufacturing” mode.It relies on Internet and cloud service platform to provide the customers with the services they need.The implementation of cloud manufacturing service can be well reflected by allocating resources properly and thus improving effective utilization ratio of manufacturing service and efficiency.At present,many researches on the match of cloud manufacturing service resources have been made.But the drawbacks of low efficiency,long matching time,lacking matching conditions and poor global optimal adaptability still exist.In this way,it is hard to solve multi-objective searching and matching problems.Therefore,an optimization method of cloud manufacturing match relying on the features of cloud manufacturing service resources and quality of service(QoS)evaluation index is proposed.Main research work can be seen as follow.Firstly,based on the whole process of cloud manufacturing services match,a corresponding optimized matching model is created to describe and analyze a certain practical problem mathematically.This model makes it possible for the key technical problems of resources match under the mode of cloud manufacturing to be discussed.Secondly,Pareto optimization is used as the key way to solve multi-objective problems.Based on the cloud manufacturing QoS evaluation system and cloud manufacturing service match optimization model,a multi-objective match model is designed.Also,the differential evolution and self-adaptive hybridization probability are combined with common bees algorithm to form a new type of algorithm,which is called ADEBA-Pareto.In this algorithm,neighborhood searching step is improved in the bees algorithm to enhance the ability of convergenceandavoid premature convergence,the match quality and efficiency would be enhanced.Optimal solution could be provided by using modified algorithm to create Pareto optimal set.With the optimal set,the most proper service would be recommended to the customers considering their preference,which could help avoid the inaccuracy brought by a mere linear weighting method in the multi-objective problem.Thirdly,according to the features of cloud manufacturing services,load balance is considered in the final stage.Threshold value is used to verify whether two manufacturing services share similar service information.If two services could satisfy the requirements of a task at the same time,the service with lower load would be recommended.ADEBA-Pareto algorithm realizes the distribution of adjustment solutions,which could not only guarantee the quality of match,but also help allocate service resources properly in practical.In the end,the advantages of our algorithm can be shown through the validity test and simulation.Lastly,a cloud manufacturing match system is designed and made,which involves service information management module,service match module and service recommendation module.Corresponding function modules are used to set related parameters and achieve the optimal match under the requests of customers.The system can show the results under practical applications.