Development of Human Daily Behavior Model and It’s Application

In this paper, we will discuss the problem how human behavior at home could be detected and how its model could be constructed. Understanding human behavior or action is important to establish a new technology that focuses on individual behaviors. Especially, it is necessary to know human habitual behavior in order to improve life environment. In actual environment, many electric appliances, e.g. TV set, washing machine, microwave stove, used at home. In this study, in order to find human habitual behavior, a sensor for electric appliance is used provided that there exist some relations between turning on/off those electric appliances and our action. Using primitive information of turning on/off, we describe human behavior at home by a binary matrix and uses neural network to express human state at home. The simulation results show that the proposed method has good performance to detect human irregular behavior at home. Copyright c ©2005 IFAC