Research And Implementation Of Semantic-Based Home Multi-Agent Cooperative Decision Mechanism

With the rapid development of the smart home and industrial Internet industry,and the wide application of various intelligent devices,the comfort of the home and the living standards of people are constantly improving,and modern society has been inseparable from artificial intelligence.The convenience that artificial intelligence provides to people is obvious to all,but it also has a series of problems: A variety of sensors in smart homes generate a large amount of sensor data,which makes data redundancy difficult to manage,and various heterogeneous problems occur among intelligent devices,such as data heterogeneity and semantic heterogeneity.The primary problem is data heterogeneity among intelligent devices,which results in the inability to share information among devices.The second problem is that most intelligent devices work independently and cannot operate with each other to achieve collaboration.In addition,the tasks in the smart home are complicated and there is no unified management,resulting in high energy consumption.The main work of this thesis is as follows:1.Aiming at the heterogeneous problems in the smart home for Internet of things,the Web ontology language(OWL)is used to make a unified semantic description of the equipment,device coding and data coding in the IoT smart home system,based on two standards: "Smart home for Internet of things device description method" and "Smart home for Internet of things data and device coding".Then,based on the characteristics of distributed and loosely coupled environment of smart home system,TopBraid Composer and protégé are used to build the ontology of smart home multi-agent combined with multi-agent technology.2.Aiming at the problem of ontology instances consistency in the process of ontology construction and ontology updating,the consistency detection method based on HermiT Reasoner and FaCT++ Reasoner is studied and implemented.When constructing the ontology of smart home multi-agent and task knowledge base,the method can be used to detect the conflicting semantic problems of the ontology instances.3.Aiming at the problems of complex tasks,complicated operations and lack of unified management in the smart home for Internet of things,OWL is used to build home-task knowledge base to semantically describe tasks,equipment status and environmental status in smart home.The various control strategies in the smart home for Internet of things are transformed into inference rules described by RDF Prolog and Lisp.The rule base is built first and then called in task collaboration.4.Semantic Web technology is introduced for the problem that the cooperation among agents in the smart home for Internet of things is poor and cannot interoperate.Combined with the multi-agent collaboration mechanism,a semantic-based collaborative decision-making mechanism for multi-agent in the home is designed to plan complex tasks.At the same time,the task conflicts in the system are analyzed to realize the detection of conflicts and the resolution of conflicts.Through the cooperation among intelligent devices,the function of making corresponding decisions on the external environment is finally realized.5.A semantic-based collaborative decision-making mechanism for home multi-agent is implemented,and a test platform is built to test and verify the system functions.The test results show that the mechanism has the ontology instance generation function,the ontology instance mapping function and the query early warning function.At the same time,the task planning decision mechanism based on the semantic Web technology is tested and verified.The results show that the platform has the ability of multi-agent collaborative decision-making in the smart home system.