| Summary: | 博士 === 國立成功大學 === 工程科學系碩博士班 === 101 === As the raising of awareness about energy crisis, the development of intelligent energy saving system becomes a new trend in all areas. In the smart grid domain, home energy management systems (HEMS) can provide energy information for residents to realize actual consuming periods, utilities, and appliances. However, because of various appliances and different spending habits of residents, general HEMS suffer from the lack of dynamic appliance recognition and bidirectional commanding. This study designs an intelligent HEMS which leverages an intermediate management layer to enhance appliance integration and reveal service API to application layer. The deployment of HEMS usually requires installation of sensing devices and specific appliances. For improving system adaptability and stability, according to the Internet of Things (IoT) architecture, this study reconstructs the platform and schemes a process of multi-appliance recognition with smart sockets. In addition, at home, human activities are probably relevant to appliance usage. Inferring activities through appliance recognition can enable the system to precisely supplies residents with electrical information and energy plans. This study provides a design of home activity inference in the system to achieve the capability of reminding residents of the appliances no longer used. In accordance with analysis of experiments, the designed system not only possesses high practicability, but improves deficiencies of other management systems.
|
|---|
