Research On Cooperative Evolution Mechanism In Manufacturing Collaborative Network

With the rapid development of network and information technology,a new service-oriented networked manufacturing model--cloud manufacturing came into being.In an open networked manufacturing environment,the different types of manufacturing resources are encapsulated as independent,autonomous manufacturing services that can dynamically aggregate on demand and cooperatively manufacture to complete manufacturing tasks.The establishment of a suitable cooperative relationship among manufacturing services is an important cornerstone for the smooth implementation of the cloud manufacturing.Regarding manufacturing services and manufacturing tasks as two types of the nodes,the dynamic connection among them is described as edges,and the cloud manufacturing can be described as a manufacturing collaborative network with dichotomy.The manufacturing tasks allocation and manufacturing services matching will change the structure of the manufacturing collaborative network,and the evolution of the manufacturing collaborative network structure will have an effect on the behaviors of manufacturing services.At the same time,the behaviors of manufacturing services will be abrupt due to the sense of injustice and the influence of the external environment.The technology combing the evolutionary game theory with the multi-agent modeling is put forward in this paper to deeply research the cooperative evolution mechanism among manufacturing services in manufacturing collaborative network.The main research contents are as follows:(1)The construction of the dynamic evolution model for manufacturing collaborative network.The characteristics of multiple manufacturing services completing tasks in the manufacturing collaborative network are analyzed,and a service quality evaluation system for manufacturing collaborative network is built.Considering that factors such as manufacturing cost and execution efficiency,the resource information understood by tasks is imperfect.When manufacturing services match tasks,tasks adopt the local preferential selection strategy and the global preferential selection strategy to select services.Based on different preferential selection strategies,different dynamic evolution models of the manufacturing collaborative networks are modeled.(2)Research on the influence of manufacturing collaborative network structure on the evolution of the cooperative behaviors.The unweighted projection method and the dynamic weighted projection method of manufacturing collaborative network are proposed,and the dynamic evolution model of manufacturing collaborative network is analyzed by projection analysis.Combining the evolutionary game theory to analyze the behaviors among service nodes,the behavior learning strategy among service nodes suitable for the manufacturing collaborative network is proposed and the behavior model of service nodes is established.Finally,the multi-agent simulation tool is used to build the service node behavior evolution simulation platform,and the experiments are carried out to study the impact of the manufacturing collaborative network structure on the cooperative behaviors evolution among service nodes.(3)Research on the impact of behavior mutation on the cooperative evolution.Combining evolutionary game theory and fair utility function,the behaviors among service nodes are analyzed.The behavior learning strategy considering the incomplete information characteristics of service node is proposed.Aiming at the mutation behavior of service nodes,the cusp catastrophe model of service node's behavior is established and the behavior mutation model is analyzed through empirical research.Finally,based on the simulation platform,the experiments are carried out to study the impact of the behavior mutation of service nodes on the cooperative evolution under the evolutionary stability of the manufacturing collaborative network.