Optimal Allocation Method Of Manufacturing Resources Based On Fuzzy Clustering And Genetic Algorithm

In recent years, the emergence of advanced networked manufacturing modes, such as cloud manufacturing, has promoted the manufacturing industry towards globalization and integration. In cloud mode, the optimal allocation of manufacturing resource is the key to realize the concept of"manufacturing is service". The study of service encapsulation, clustering analysis and optimal allocation of manufacturing resource in this thesis has important practical significance for improving service efficiency and large-scale popularization of cloud manufacturing.The service description model and virtualization encapsulation mechanism of manufacturing resource service were constructed. With the characteristics such as mass, heterogeneity and independence in manufacturing resource, the service description model of manufacturing resource was established and the functions and significance of each module in model were discussed based on the object-oriented thought. For the sort of verifying problem of resource structure and semantics, one kind of manufacturing resource virtualization encapsulation framework was built on the basis of Web.Fuzzy clustering research of manufacturing resource was done. Because of the problems of poor correlation relation, unfavorable realization of high-efficient retrieval and optimal allocation among manufacturing resources, direct clustering method based on the fuzzy similarity matrix and FCM clustering method were adopted respectively to cluster and analyze one certain batch of manufacturing resources. Simultaneously, problems such as setting attributes’ weight value and choosing clustering center were discussed and improved, and one new method combining the above two clustering methods together to use was proposed.Optimal allocation algorithm of manufacturing resource was improved. According to the optimal allocation problems of manufacturing resource in the cloud mode, the corresponding evaluation function model was established, the traditional genetic algorithm was improved adaptively and the selection efficiency of the algorithm was improved by using the combination roulette and tournament selection. The diversity and convergence of algorithm were guaranteed by adopting adaptive crossover rate and mutation rate, and the stability and solution quality of the algorithm was raised by providing positive feedback. Finally, the improving algorithm’s effectivity was examined and the optimal resource allocation solution was produced.An optimal allocation prototype system of manufacturing resources was designed and implemented. According to many requirements in manufacturing resource allocation research, with cloud manufacturing service platform as an study object, overall structure of the platform,functions of each module, theory and technology were designed and developed, which verified the feasibility and effectivity of the theory and method studied in this thesis.