Social Implications of Cyber-Physical Systems in Electrical Load Forecasting

With the further development of smart grids, lots of communication devices such as sensors and actuators are dramatically equipped into power grids. At the same time, social behaviors, investment, trading, management and user selection have become increasingly important in energy system research. The modern power system has evolved from the Cyber-Physical-Systems (CPS) which combines the power network and the cyber network to the Cyber-Physical-Social Systems (CPSS) composed of the cyber network, the power network and social network. For CPSS system, how to reflect its social characteristics is one of the biggest difficulties. Based on the social policy data, this paper constructs a statistical model reflecting the load law of power system under the framework of social physical information system, then it uses the improved long and short-term memory (Long Short-Term Memory, LSTM) deep learning network to train the model. Finally this paper realizes the integration drive based on data, model and social factors.