Abstract

Under the background of “Carbon peaking, Carbon neutralization”, in order to effectively control the future carbon emissions, it is necessary to accurately and quickly predict the regional energy consumption in the short term. Therefore, this paper proposed an energy consumption prediction method based on VMD-LSTM. In order to solve the problem of large change rate of short-term energy consumption and noise in the consumption curve, the energy consumption curve was decomposed into multiple intrinsic mode functions (IMFs) and reconstructed by variational mode decomposition (VMD). The reconstructed curve was evaluated by noise reduction evaluation index. The energy consumption curve after noise reduction was forecasted by long-short term memory network (LSTM) and updated the network by real value in each step. The results of simulation that forecasted the power consumption showed that the energy consumption prediction method based on VMD-LSTM can accurately predict the short-term energy consumption with shorter time, and can provide guidance for the effective planning and control of carbon emissions.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call