Learning and adaptive fuzzy control system for smart home

Abstract

Automated smart homes have widely established their position as a research field during the last decade. More and more context sensitive concepts are being studied and at the same time proactivity has broken through in ambient intelligence research. Technology has advanced towards an adaptive and autonomous home, which can take care of the inhabitants' well-being in numerous ways. We propose to use a context sensitive and proactive fuzzy control system for controlling the home environment. Since humans communicate using fuzzy variables, we see that applying context recognition to a fuzzy control system is straightforward. The designed control system is adaptive, and it can accommodate to changing conditions of inhabitants. Our system is designed to operate fully in the background and needs very little effort from its users. The system utilizes a principle of continuous learning so that it does not require any training prior to use. This paper describes a lighting control system implemented using the fuzzy control system designed. We concentrate on the basic operation of such systems and present findings from the design process and initial tests. © 2006 Springer-Verlag France, Paris.