Font Size: a A A

Research On The Involving Collaborative Mechanisms In Manufacturing Collaborative Network Based On Multi-Agent

Posted on:2019-03-15Degree:MasterType:Thesis
Country:ChinaCandidate:J B WangFull Text:PDF
GTID:2428330596965425Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
The paradigm of service-oriented networked manufacturing,the cloud manufacturing provides a new idea and direction for the development of the manufacturing.In the cloud manufacturing platform based on the open networks,the massive,distributed and dynamic manufacturing resource of various type are encapsulated into manufacturing services to complete the manufacturing tasks through cooperation to attain the manufacturing of tasks on demand.The establishment of a cooperative relationship among the manufacturing services is one of the important foundations for the realization of the cloud manufacturing.Hence,it is necessary to study the evolution mechanism among the manufacturing resources.If the manufacturing tasks and the manufacturing services are respectively treated as two types of the nodes and the relationships among them are expressed as the edges,the cloud manufacturing can be represented as the bipartite network---manufacturing collaborative network.Combined with the theory of the evolutionary game,the research on the cooperation relationship among the manufacturing services is essentially based on the study of the game relationship among the manufacturing services.Hence,combining the technology of multi-agent modeling,the paper studies the cooperation and the evolution mechanism among the manufacturing services.The main research work is as follows:(1)Based on the background of the cloud manufacturing,the characteristics of manufacturing network are analyzed to build the relationship network among the service nodes based on the evolutionary models of the manufacturing collaborative network and combined the technology of the projection of the bipartite network.Then,based on the evolutionary game theory,the game relations among the service nodes is analyzed and the game theory of the service nodes are proposed.Finally,based on the game model and the evolutionary model of the manufacturing collaborative network and combined the technology of the multi-agent modeling,the service nodes are encapsulated as agent to establish the behavior model of the service nodes.(2)According to the strategy learning mechanism in evolutionary game theory,based on the background of cloud manufacturing and the characteristics of manufacturing collaborative network,three different strategy learning mechanisms are proposed in the manufacturing collaborative networks.The mechanisms includes the self-learning mechanism based on Q learning,the strategy imitation mechanism in the relationship network among the service node and the mixed imitation mechanism based on the former two.The principle of the three mechanisms is used to analyze and predict the influence of different strategy rules on the evolutionary cooperation relationship of service nodes.(3)Based on the multi-agent modeling technology,the paper uses the simulation tool to establish a cooperative evolutionary simulation platform for manufacturing collaborative network for a typical cloud manufacturing instance..Based on the experiments of the simulation platform designing,the paper studies by mathematical tools to study the behavior evolution of service nodes under the important model parameters and under different strategic mechanisms.Finally,the simulation data is processed by the mathematical tool,the result is obtained after analyzing the data.The result shows that the development of collaborative environment between service nodes in manufacturing collaboration networks relies on effective cooperation incentive mechanisms,especially in the case of that individual nodes tend to refer to the behaviors of other individuals.
Keywords/Search Tags:Manufacturing Collaborative Network, Manufacturing Service, Evolutionary Game Theory, Strategy Learning Mechanism, Multi-agent Modeling
PDF Full Text Request
Related items