Development of Hour Ahead Demand Response Algorithm for Smart Home

The optimization of energy consumption with a consequent cost reduction is one of the struggles in demand response under the Smart Grid environment. Under the DR paradigm, consumers are expected to change their consumption behavior by shifting appliances from on-peak period to off peak period. In this paper, Artificial Neural Network based Demand Response (DR) algorithm is proposed to predict the hour ahead price of the energy supplied from the utility. Time, day, holiday or working day, historical power demands and historical electricity costs are considered inputs and predicated price is the output of ANN. The hardware realization of DR environment is developed with Raspberry Pi. In which, Python based ANN is developed and different kinds of loads are interfaced with controller. Blynk app is utilized to provide information on demand level during off-peak, low-peak, and on-peak period.