Research Of Intelligent Information Appliances Based On Multi-Agent

The traditional electronical appliances have been unable to meet the needs of the modern home, people need the information appliances which are able to automatically acquire outside information, cooperate with other information appliances with intelligence, so the development of intelligent information appliances is the inevitable trend of the future home appliances development.It is of very important theoretical and practical significance to take advantage of the Agent initiative, intelligence, interactivity, collaborative learning, mobility, etc. and apply the intelligence Agent technology to the research of information appliances. This also provides the new way of thinking for the research. In the future home appliances systems, there are different types of household electric appliance. MAS provide a mechanism which effectively connect the separate and independent household information electric appliances and complete complicated tasks through the coordination, and also provides a relatively loose control and relatively flexible for entry and exit of the single information appliances. This is of great significance for the information appliances environment in the process of dynamic change, which could hardly be realized in the traditional concentration control mode.Based on the need of the information appliances intelligence trend, this article considers every information appliance as an intelligent unit. These intelligent units could automatically set their own parameters and act initiatively, and proposes the information appliance mode—MIAS based on the MAS technology. It also makes a research on every part of the mode, summarizes the features and types of every Agent in the mode. At the same time, this article analyzes various types of Agent concept mode according to the relative theories about the Agent, offers the realistic modes of the information appliances Agent, family gateway Agent and organization planning Agent. It designs a function module with great practicality, whose structure can stand for the telecommunication of the IA equipment and reason and plan the behaviors. The structure also provides a uniform platform to explain the automation of the IA Agent and construct different types of Agent. Through the system completeness, sequence realized from the coordination of every information appliance, in view of the MIAS with many Agent learning requirements, and on the basis of the learning intensification, and combined Q-Learning and the advantages of the fuzzy logic, the article proposes the fuzzy-Q learning algorithm, makes the IA Agent autonomous intelligent reasoning base on the data calculation, thus increase the expression and reasoning ability with IA Agent knowledge.