Research On Personalized Rule Generation And Application Of Multi-agent Technology In Smart Home

Advances in smart devices, mobile wireless communications, sensor networks, pervasive computing, machine learning, middleware and agent technologies, and human computer interfaces have made the dream of smart environments a reality. Pervasive Computing is considered as a new computing model following the mainframe computing model and desktop computing model.In Pervasive Computing environment, people can utilize computing services whenever and wherever. In fact, pervasive computing in the physical space and information space can be fused in varying degrees, but also can reflect on them. Pervasive computing applications can be embodied n a room or in a building. If you are in a room or in a building to achieve the physical space, which can be claimed as a smart space.Since the development of pervasive computing, based on ubiquitous computing smart home space systems have emerged. Smart Home space is a distributed multi-Agent system. Various hardware devices had been embedded in the smart home. The user's information at any time by embedded devices with access to and acquisition. Users can also use the services of home devices, but such an intelligent family of multi-Agent system, users must use the resources to predict, so to avoid resource use conflicts. However, because the use of the resources each user is different. Personal preferences are not the same. Users are not the same for controlling their information. Therefore, the user is in the family tried to formulate some rules to their own information processing, these rules not only to Agent for the user but can predict the use of resources, while users effective protection of information in order to facilitate a better family of intelligent for customer service. This paper first studies the problem of generating the individual rules and rule generation of the dynamic resource constraints, and then introduces the semantic rules of inference model; finally, a rule inference verification. The main work contents and innovations in this paper are summarized such as:On the formulation of rules of individual users, this paper first proposed strategy to generate the basic framework, which detailed description of the smart home system, rule generation process. In the generation process, the use of a collaborative filtering algorithm to calculate the cosine vector feature similarity and type of degree of similarity to get to the user as an integrated similarity task Agent by using machine learning algorithm to generate the user recommended strategy. In this way, users recommend strategies based on system settings, which greatly improves the system policy settings for speed and accuracy.In the smart home, due to limited resources, in order that resources are effectively utility. According to the rules set Agent for each user dynamically generated for each user of a resource constraint rules. This multi-Agent system for the prediction of a NSL learning framework presented so that every user at the same time to achieve a satisfactory result, that is, to achieve a balanced effect, that is, Nash equilibrium. The algorithm also provides the basis for the regulation of resources.Based on individual rules and Resources constraint rules, the question that introduced the formal definition of semantic rules, and describes the rules in JESS machine representation, and then reasoning model is given.Finally, the article supporting platform system design model was given, and the rules set by the user and dynamic resource prediction set to validate the model's practicality and effectiveness, and then a summary and outlook for the next step for further research to lay a solid basis.