In view of the current problem of lack of key guiding signals for low-carbon energy consumption on the user side, a dynamic carbon emission factor prediction method taking into account the low-carbon demand response behavior on the user side is proposed. First, a user dynamic electricity carbon emission factor calculation model is constructed based on the carbon emission flow theory, and a carbon emission factor data pool is constructed in combination with system operation simulation. Second, a low-carbon energy consumption response behavior model for power users facing dynamic carbon emission factors is constructed, and a dynamic carbon emission factor prediction method taking into account the low-carbon demand response behavior on the user side is proposed. Carbon emission factor prediction is carried out based on LSTM neural network, and effective prediction of node-level dynamic carbon emission factors for a given system based on arbitrary source and load input is achieved. Finally, a case analysis is carried out based on a PJM-5 node power system and a 36-node power system with a high proportion of renewable energy, which verifies the effectiveness of the proposed method in predicting node-level electricity carbon emission factors taking into account the user's low-carbon demand response.