Abstract:

According to the attributes and behavior character of the entities composing the intelligent physical world brought by pervasive computing, the perception level of the KRA (knowledge reconfiguration and abstraction) model proposed by Saitta and Zucker to represent the static physical world was divided into primary perception and perception reconstruction. This can represent and distinguish the proposed entities of various domains to capture the hierarchical model of the physical world in intelligent environment. The presented extended model defines physical entities unitedly and represents networked entities logically to avoid reduplicate perception. The relationships between networked entities and between virtual entity and networked entity are logically defined to help the hierarchy of the model. In addition, the reasoning procedure based on the whole represented world can be decomposed into less\|scale ones based on the
 space constructed by physical entities and their relationships and by virtual entities and their relationships, so that the complexity of reasoning is greatly
reduced.

Key words: knowledge reconfiguration and abstraction model, intelligent environment, entity domain, networked entity, virtual entity