Comparative analysis of two decomposition methods for urban domestic water usage patterns

ABSTRACT This study presents a method for analyzing urban water usage patterns using variational modal decomposition (VMD) with K-value optimization, which can improve the understanding of short-term urban water demand fluctuations. This method is compared to the fast Fourier transform (FFT) to evaluate its effectiveness in identifying periodic, daily, and weekly water usage patterns. VMD decomposes the original signal into multiple intrinsic mode functions. FFT is applied to analyze and reconstruct the different frequency components of the signal, transforming them back into the time domain. While FFT shows greater computational efficiency and higher reconstruction accuracy, VMD excels in handling nonlinear and non-stationary signals with better decomposition capabilities. However, VMD requires more computation time, and its performance is sensitive to parameter selection, indicating room for further optimization. Both methods successfully detect the main frequencies in short-term urban water usage patterns, particularly identifying cycles of 0.5 and 1 days. VMD demonstrates significant potential for capturing low-frequency fluctuations in urban water usage, making it a valuable tool for future research and prediction of water usage trends, particularly in short-term forecasts.