Home-legacy device intelligent control using ANFIS with data regeneration and resampling

In recent years, the research for electric power usage reduction in a house has been widely studied. Home energy management system (HEMS) is become one of major applications to handle many smart electrical devices inside house that proves its electricity usage reduction efficiently. However, HEMS has a critical issue which cannot control non-smart devices at home. Saving unnecessary energy usage for legacy devices remains research area to prevent from expansions of energy waste for users. In this paper, an intelligence inference control approach based on the adaptive neural-fuzzy inference system (ANFIS) is proposed for legacy devices. The approach based on ANFIS focuses to reduce computation of training time by performing regeneration and resampling approach compared to conventional ANFIS.