A short-term energy prediction system based on edge computing for smart city

The development of Internet of Things technologies has provided potential for real-time monitoring and control of environment in smart cities. In the field of energy management, energy prediction can be carried out by sensing and analyzing dynamic environmental information of the energy consumption side, and provide decision support for energy production to avoid excess or insufficient energy supply and achieve agile production. However, due to the complexity and diversity of the IoT data, it is difficult to build an efficient energy prediction system that reflects the dynamics of the IoT environment. To address this problem, a short-term energy prediction system based on edge computing architecture is proposed, in which data acquisition, data processing and regression prediction are distributed in sensing nodes, routing nodes and central server respectively. Semantics and stream processing techniques are utilized to support efficient IoT data acquisition and processing. In addition, an online deep neural network model adapted to the characteristics of IoT data is implemented for energy prediction. A real-world case study of energy prediction in a regional energy system is given to verify the feasibility and efficiency of our system. The results show that the system can provide support for real-time energy prediction with high precision in a promising way. (C) 2019 Elsevier B.V. All rights reserved.