Residential energy consumption and price forecasting in smart homes based on the internet of energy

Smart home applications require automatic energy management and control of unwanted energy consumption prices using Artificial intelligence techniques. Previously, there were more studies on price forecasting. However, efficient results are still under research. This study introduces real-time residential energy management systems based on the Internet of Energy with a scheduling strategy. The proposed method employs a Gated Recurrent Unit and Bird Swarm Optimizer (GRU-BSO) along with Real-Time Electricity Scheduling (RTES) based on Price Forecasting (PF) for efficient residential energy management. The research emphasises the optimisation algorithm's ability to minimise energy costs and promote energy conservation, significantly contributing to the field. Price forecasting (PF) is the central objective in distributed energy production. By forecasting the optimal price, this approach can improve the efficiency of power grids and solve issues with microgrid management and planning. It is suggested that the tariffs for shoulder-peak and on-peak hours be determined using the Time of Use (ToU) model. The proposed method also predicts the energy price used in home energy management. The cloud server and MATLAB-implemented microgrid system are linked via a two-level communications network. The current communications level uses the local communication level as a protocol, which uses IP/TCP and MQTT. The study's proposed scheduling controller successfully achieved energy savings of 17 kW and 47 cents by utilising the proposed method.