| Plate metal is widely used as an important raw material in the fields of automobile manufacturing,construction machinery,and pressure vessel manufacturing.With the development of manufacturing informatization and personalized demand,the production mode of manufacturing enterprises is gradually moving towards small batches and various types.But for small and medium-sized enterprises that are at the bottom of the product manufacturing cycle,Due to the transformation of production mode and the shortage of enterprise production intelligence and informatization,In the processing of the plate,it is inevitable produce remnant plate that will cause huge waste.However,with the integration of information technology such as big data,cloud computing,and Internet of Things and advanced manufacturing technologies,cloud manufacturing came into being.Cloud manufacturing is a service-oriented new networked manufacturing model,Its "manufacturing as a service" idea provides ideas for the effective use of remnant plate.Relying on the cloud manufacturing model,the remnant plate can be used as an important manufacturing resource to re-enter the life cycle of product manufacturing.At the same time,it can also drive the idle equipment resources of the enterprise.How to realize the service and efficient sharing of manufacturing resources of sheet metal processing enterprises has put forward urgent needs.This paper research the above problems from three aspects: manufacturing resource service,service organization and service matching.First,in order to achieve effective expression of resource information under the cloud manufacturing platform,In this paper,the characteristics of the remnant plate are analyzed,and it is classified based on the shape feature.At the same time,a layered and modular resource description model is proposed which provides a basis for resource service.In addition,modeling of remnant plate and processing equipment resources is carried out based on ontology theory.Secondly,in order to realize the intelligent classification and semantic annotation of the shape characteristics of the remnant plate,This paper relies on computer image processing technology and machine learning algorithm.The effects of FD and Hu moments on the recognition effect of the shape of the remnant plate are compared by contrast experiments.Compare the influence of different length FD on recognition accuracy.FD were used as feature vectors to compare the effects of five different supervised classification algorithms on the recognition accuracy of remnant plate.Thirdly,using ontology advantage,an ontology semantic similarity matching algorithm is proposed for remnant plate processing enterprises.The algorithm is divided into three stages: remnant plate primary selection,equipment resource matching and QoS matching.Convert resource services one by one into hierarchical,phased matching to improve matching efficiency.By analyzing the ambiguity of QoS indicators and transforming them into triangular fuzzy numbers,Using user preferences,a fuzzy QoS matching method using matrix features is proposed.Finally,the matching method of this paper is verified by an example. |
