Modeling of Smart Energy Cities Based on Deep Neural Networks

Aiming at the smart energy city currently has a wide range of energy data sources, including electricity, natural gas, heat and many other different fields of energy data often have differences in format, standards and collection methods, making it difficult to achieve uniform integration and analysis problems. Based on deep neural network technology, this paper proposes a smart energy city model, aiming to construct a city physical examination index system under the energy perspective. Firstly, by integrating energy data from different fields such as electricity, gas and heat, and combining the three key elements of “people, land and housing”, a comprehensive and systematic city checkup evaluation model is constructed based on deep neural network. Secondly, on the basis of the existing city checkup evaluation system, the metadata, dimensions and calculation methods of each index are clarified to analyze the close relationship between energy use and urban development. The study shows that constructing a city checkup indicator system from the energy perspective can not only comprehensively reflect the city's energy utilization situation, but also promote the optimization of energy structure, reduce energy consumption, provide decision-making support for green and low-carbon development, and enhance the level of city management.