Research On Game Theory Based Real-time Scheduling Strategy And Method For Flexible Job Shop In The Internet Of Manufacturing Things

With the continuous integration of information technologies such as the Internet of things(Io T),cloud computing and big data with the manufacturing industry,a new manufacturing mode-the Internet of Manufacturing Things(Io MT)has emerged.In the Io MT environment,the manufacturing enterprises are changing their production modes from the “black-box” mode to “the multi-dimensional and transparent ubiquitous” mode,which proposes a higher requirement for real-time monitoring and dynamic optimization of the production process.Meanwhile,as the key sector to ensure highly efficient and quality of production process,workshop scheduling has been widely concerned and studied,especially in how to improve the dynamic response capability and real-time optimization performance of workshop scheduling.Under this background,this study mainly focuses on the key technologies of “Game theory based strategy and method for flexible job shop real-time scheduling(FJSRS)”,including the overall architecture for the FJSRS based on game theory,the static game based single-objective FJSRS,the dynamic game based multi-objective FJSRS and the infinitely repeated game based multi-objective FJSRS considering dynamic changes of production planning.By introducing Io T technologies to the real-time scheduling,it could capture the real-time information during the process real-time scheduling,promote real-time optimization allocation of production tasks and efficiently improve the manufacturing resource management level of manufacturing enterprises.The main works of this study are summarized as follows.1.An overall architecture of the game theory based FJSRS is proposed in the Io MT environment.Firstly,the definition and classification of FJSRS are described,and the relevant concepts of Io MT are introduced.Then,the advantages of FJSRS strategy compared with traditional dynamic scheduling strategy are analyzed.Next,based on the game theory,the FJSRS architecture for dynamic response and real-time optimization of production tasks under the Io MT environment is proposed and described in detail.Finally,the key technologies are analyzed.2.A static game based single-objective FJSRS method(SGSO-FM)is proposed for single-objective FJSRS problem.Firstly,a mathematical model of single-objective FJSRS is established to maximize the average utilization of the machine.Then,a single-objective FJSRS strategy is proposed.Next,the static game based single-objective FJSRS model is established from the point of view of the static game.Furthermore,the static game method for solving the single-objective FJSRS and the algorithm for solving the Nash Equilibrium(NE)are described in detail.Finally,the validity and feasibility of SGSO-FM are verified through a simulation experiment.3.A dynamic game based multi-objective FJSRS method(DGMO-FM)is proposed for multi-objective FJSRS problem.Firstly,a mathematical model of multi-objective FJSRS is established to minimize the maximum completion time,minimize the total load of the machine and minimize the total energy consumption of production.Then,a multi-objective FJSRS strategy is proposed.Next,considering the decision sequence of participants,the dynamic game based multi-objective FJSRS model is established from the point of view of the dynamic game.Furthermore,the dynamic game method for solving the multi-objective FJSRS and the algorithm for solving the sub-game perfect Nash Equilibrium(SPNE)are described in detail.Finally,the validity and feasibility of DGMO-FM are verified through a simulation experiment.4.An infinitely repeated game based multi-objective FJSRS method considering dynamic changes of production planning(IRGMO-FM-DCPP)is proposed for multi-objective FJSRS problem.Firstly,a mathematical model of multi-objective FJSRS and the infinitely repeated game based multi-objective FJSRS model is established considering dynamic changes of production planning.Then,the infinitely repeated game method for solving the multi-objective FJSRS considering dynamic changes of production planning and the algorithm for solving the SPNE are described in detail.Finally,the validity and feasibility of IRGMO-FM-DCPP are verified through a simulation experiment.5.Based on the above key technologies,a simulation system is designed and a case is verified.Firstly,the FJSRS simulation system is developed under the JADE(Java Agent Development Framework)platform.Then,the operation processes of single-objective FJSRS,the multi-objective FJSRS,and the multi-objective FJSRS considering dynamic changes of production planning are described respectively.Finally,an industrial case is discussed to verify the function of the simulation system.The results show that the proposed theory,models,strategies and methods are feasible.