Construction Of Service Coalition In Collaborative Manufacturing Environment

As a new manufacturing mode,collaborative manufacturing has the advantages of shorter production cycle,faster response,and lower production cost than traditional production scheduling modes.However,due to the distribution and heterogeneity of resources,collaborative manufacturing faces the following problems: lack of collaborative service capabilities of enterprises and crisis of trust among enterprises.These problems will restrict the long-term development of enterprises,it also will lead to unnecessary waste of corporate resources and undermine the cooperation between enterprises.The thesis focuses on the above issues and the specific work is as follows:Firstly,this thesis proposes a trust-guided service coalition model.This model studies from the perspective of users,coalition and coalition members,accurately describes the functions of various modules within the service coalition,and the trust relationship between the members.Based on this model,a trust-guided service coalition algorithm is proposed.The service coalition resource scheduling is optimized with the goal of maximizing the revenues and the highest integrated trust value of the coalition service portfolio.Secondly,this thesis gives a dynamic update method of comprehensive trust value.First,the comprehensive trust of the coalition members are calculated based on system trust,recommendation trust,and historical trust.Then it decomposes the integrated feedback trust values of the service portfolio by users and coalition to each member,so the trust value of each coalition member is updated dynamically.Lastly it makes system of rewards and penalties to ensure the superiority of the coalition members.Finally,this thesis constructs a mathematical model of coalition members selection.In this model,the members selection variables are transformed into decision variables in the evolutionary game algorithm;The objective function fitted to the three indicators of unit service price,integrated trust value,and response time of the members are transformed into the revenue function in the algorithm;The process of finding the best member is transformed into the process in which the game participant searches for the optimal Nash equilibrium in the game to obtain the best member.Thesimulation experiments show that the trust-driven evolutionary game algorithm is more effective in the selection of the member selection problem.