Cooling load prediction of air-conditioning based on VMD-TCN using PE-SG algorithm

Accurate prediction of air-conditioning cooling load is not only beneficial to control energy consumption and improve energy efficiency, but also provides a theoretical basis and data support for energy conservation and emission reduction. Aiming at the problems that large deviation of original data and low prediction accuracy in air-conditioning cooling load prediction. An air-conditioning cooling load prediction model combined with time convolutional network (TCN) combined with permutation entropy (PE), savitzky-golay (SG) and variational mode decomposition (VMD) is proposed in this paper. Firstly, Pearson correlation coefficient is used to analyze historical data. Secondly, the complex multi-component cooling load signal is decomposed into multiple single-component amplitudes and frequency modulation (AFM) signals by VMD. The PE is used to quantitatively determine the noise content of each component after VMD decomposition, the high noise component is directly removed, the low noise component is smoothly processed by the SG smoothing method, then, the signal is reconstructed. Finally, the TCN model of air-conditioning cooling load prediction is established. The experimental results show that the prediction accuracy of the hybrid model is significantly improved compared with the conventional models.