The intelligent smart home provides various services offering and amount of knowledge reasoning. However, programmers need to consider about constraints including scenarios in different houses, scenarios for different users, and even different resources. That is, infrastructures deployment and scenarios design are truly time consuming. Actually, it is more reasonable to make a home adapt to family members' behaviors than a user adapt to limitation functionalities of a home. Therefore, development of an efficient reasoning system for smart home is widely discussed in recent research. In this paper, we propose a smart home reasoning system called ASBR system. The system learns user's preferences by adaptive history scenarios and offers a convenient way to rebuild reasoned knowledge in other smart homes. We first introduce that ontology based context information in smart home can be extracted and reasoned as a set of scenarios. In addition, system can derive personalize habits and store into OWL files. We then present scenario reconstruction method under computation and resources restrictions. Finally, we design an experiment in a realistic smart home and propose some scenarios to discuss work result.