Smart home will improve the quality of life by providing services actively that are adequate on a user’s situation in household, entertainment, etc. If smart home cannot guarantee the reliability up to the level expected by a user, it is almost impo...
Smart home will improve the quality of life by providing services actively that are adequate on a user’s situation in household, entertainment, etc. If smart home cannot guarantee the reliability up to the level expected by a user, it is almost impossible for smart home to be adopted due to enterprise’s spending on maintenance. Because smart home consists of various devices, complex software and various network technologies, it is not easy to guarantee the reliability. So we need a method that diagnoses and recovers smart home’s fault.
Previous researches cannot deal with various types of faults. A fault may occur from conflicted commands for an identical component and relations among commanded components other than software and devices faults. These types of faults can be identified with fault detection and diagnosis knowledge. It is difficult for a person to generate fault diagnosis knowledge or to acquire that by setting a real smart home. This thesis designs and implements FOSIM(Fault diagnOsis knowledge SIMulator) that generates fault diagnosis knowledge through smart home simulation.
The Simulator consists of SOSIM(Smart hOme SIMulator) and FOGEN(Fault diagnOsis knowledge GENerator). SOSIM simulates smart home for generating meaningful knowledge. We select necessary components for smart home such as devices, services, environments and model those using Entities. We classify theses components into environment layer, device layer, service layer and simulate components and their relationships. FOGEN generates fault diagnosis knowledge using contexts generated by Context Generator. It finds fault symptom according to fault detection knowledge and contexts, identifies reasons of faults from contexts using dependant relationships among devices, and generates the fault diagnosis knowledge.
We verified fault diagnosis rules generated by FOSIM. The longer FOSIM was executed, the more fault diagnosis rules were accurate. Smart home can diagnosis various faults using fault diagnosis rules.