In the context of "carbon peaking and carbon neutrality"，carbon emission accounting results are an important basis for the government to formulate carbon emission reduction policies.Due to the difficulty in obtaining data required by regional carbon emission calculation，traditional calculation methods depending on these data have weak applicability and low accuracy for carbon emission calculation in urban areas.To this end，a method for predicting carbon emissions and analyzing influencing factors in urban areas considering the electricity consumption in key industries is proposed.The conversion relationship of electricity，energy，and carbon emissions in the regional key enterprises，residential buildings，and transportation is explored.The carbon emissions of various industries in the region are calculated based on the carbon emission data of these enterprises.The dynamic time warping （DTW） model is used to calculate the correlation between these industries and electricity-related carbon emissions in the region，and key electricity consumption industries in the region are selected through box plots.A regional carbon emission prediction model is established based on a long short-term memory （LSTM） network，and the prediction results of regional carbon emissions are obtained.A regional carbon emission influencing factor analysis model is constructed based on stochastic impacts by regression on population affluence and technology （STIRPAT ） model.By taking the urban new district in the eastern region as an example，the carbon emissions of this new district in 2022 are predicted，and the main influencing factors of the changes in carbon emissions of the new district are analyzed.A reference is provided for carbon emission prediction in urban areas considering electricity consumption of key industries.