Balanced broad learning prediction model for carbon emissions of integrated energy systems considering distributed ground source heat pump heat storage systems and carbon capture & storage.

• Distributed ground source heat pump heat storage system is considered. • Energy storage is applied to balance the energy network of integrated energy systems. • The carbon capture and storage technology are applied to store carbon. • A balanced broad learning prediction model is established for load forecasting. • The optimal dispatch model is verified to achieve carbon neutrality under four cases. With the development of social industry, numerous carbon emissions have led to global warming and caused a huge negative impact on the environment. Carbon neutrality has been proposed to solve this problem; one of the primary issues in achieving carbon neutrality in integrated energy systems is to reduce carbon emissions. This paper utilizes distributed ground source heat pump heat storage systems in integrated energy systems to simultaneously improve the utilization efficiency of wind energy and solar energy for the first time. Moreover, the carbon capture and storage technology in integrated energy systems is considered to store excess CO 2 in geological layer. Energy storage in integrated energy systems is applied to adjust the energy network balance. Finally, a balanced broad learning prediction model considering various heterogeneous data is established for load forecasting with 96.12% prediction accuracy. In four cases of the lower or higher wind and solar energy generation curves, carbon emissions are reduced to 82.02% of the original at least. [ABSTRACT FROM AUTHOR]