| Manufacturing industry is an important industry of China's real economy.Facing the new normal requirements put forward by Made in China 2025,it is an inevitable trend to transform information into data and share media perception functions through the Internet as the development direction of transformation and upgrading.In order to break the phenomenon of manufacturing island and improve the utilization efficiency of manufacturing resources,cloud manufacturing,a new manufacturing mode,is constructed by combining the powerful data processing ability of cloud computing and information technology such as Internet of things.Cloud manufacturing enables users to obtain manufacturing life-cycle services on demand anytime and anywhere.It realizes a high degree of resource sharing and service value-added efficiency,which is the basis of promoting China from a big manufacturing country to a powerful manufacturing country.Cloud manufacturing is based on a service-oriented manufacturing cloud platform,in which service composition is one of the core points of current research.Although many scholars have carried out in-depth research on service composition and achieved some results,there are still deficiencies in how to choose a service composition that meets user needs and can form the best solution from massive cloud manufacturing services(service association not considered,single subtask and service mapping and the design of optimal method).This paper makes a further discussion.Firstly,after fully understanding the service composition process,key technologies and optimization methods in the cloud manufacturing environment,the description model of the manufacturing service and the basic evaluation model of service quality QoS are studied.The relationship between the interfaces and the business entities between the services is studied.And based on previous studies,the description model and QoS calculation method of business entity correlation and statistical correlation are improved.In the business entity correlation,the cooperation time of enterprises is considered to provide a more stable cooperation scheme.In order to make the statistical correlation more accurate and timely,the two-way confidence KULC and the imbalance factor IR are used to evaluate the association rules,and the "time-sharing effective" idea is adopted for QoS prediction,forming a dynamically adjustable QoS changes.Thus,a QoS perception evaluation model supporting service association is constructed.At the same time,crowdsourcing mode is introduced into the mapping relationship between subtasks and services,and a new set of candidate services is established through crowdsourcing service screening.Secondly,in the design of cloud manufacturing service composition optimization algorithm,sparrow search algorithm has better convergence effect,fewer parameters and simple code.However,population initialization and local optimization are the defects of the algorithm.This paper proposes a sparrow search algorithm based on vertical and horizontal crossover.On the one hand,tent chaotic map is used to generate better quality initial sparrow population to speed up the global optimization process and improve the accuracy of the solution.On the other hand,the vertical and horizontal crossover strategy in hybrid crossover optimization algorithm updates the population mutation from the vertical and horizontal dimensions to increase the optimization range and prevent the emergence of local optimization problems.Finally,an example is used to verify the effectiveness of the QoS perception evaluation model designed in this paper based on the actual manufacturing background.And the performance advantage of the improved sparrow search algorithm CSOSSA in solving the optimization problem of cloud manufacturing service composition is proved through simulation experiments. |
PDF Full Text Request