A game strategy for demand response based on load monitoring in smart grid

Abstract

Demand response technologies can achieve the objective of optimizing resource allocation and ensuring efficient operation of the smart grid by motivating the energy users to change their power usage behavior. However, the increasing uncertainty of smart grid environment brings great challenges to the development of demand response technique. In this study, we build a long short-term memory (LSTM) network as a load forecasting model to predict user load data in order to obtain accurate consumption behavior of energy users. Then, we utilize a Stackelberg game model based on the load forecasting model to dynamically optimize the electricity prices set by power suppliers at different times, enhancing the efficiency of demand response between users and suppliers. Extensive simulation experiments demonstrate that the LSTM-based load forecasting model achieves an accuracy of up to 96.37% in predicting user load demand. And the game model reduces the overall expenditure of users by 30% compared with the general pricing model.