Aquisição e processamento de sinais de vibração em redes de distribuição de água para detecção de vazamentos

Abstract

This work investigates the use of signal processing techniques for leak detection in water distribution pipelines, focusing on the application of the Fast Fourier Transform (FFT) and the Continuous Wavelet Transform (CWT). The study analyzes vibration signals captured by sensors under both leaking and non-leaking operating conditions, aiming to identify characteristic frequency patterns of each situation. The FFT was applied to obtain a global view of the frequency components, while the CWT allowed a detailed time-frequency domain analysis, revealing transient variations associated with leaks. Although the CWT offers a more comprehensive analysis, its application requires greater computational capacity, making the FFT a more viable option for continuous long-term monitoring due to its efficiency. The results show that the combination of FFT and CWT provides a robust solution for leak detection, helping to reduce water losses and promote resource sustainability. This study contributes to the development of more efficient and accurate monitoring systems for the water utility sector.