Classification of home electrical appliances using deep learning for smart grid

One of the basic conditions for the successful implementation of energy demand side management (TTY) in smart grids is the monitoring of different loads with an electrical load monitoring system. Energy and sustainability issues present a multitude of issues that can be addressed using data mining and machine learning approaches. However, resolution of such problems has been slow due to the lack of publicly available datasets. In this study, first an efficient Energy Disaggregation model was designed and model evaluation was performed based on publicly available benchmark data and success metrics called the Residential Energy Disaggregation Dataset (REDD). A dataset we call the Turkish Electrical Appliances Dataset (TEAD), which includes data on domestic and household electricity usage, is presented. In addition, in this study, the device classifier model proposed with TEAD was trained and performance evaluation was made. The success of this evaluation was confirmed by the performance test with the REDD data set. In this thesis, Internet of Things (IoT) architecture was established with current sensors and Node-Red software installations to collect data.