Leakage detection of HVAC pipeline network based on pressure signal diagnosis

In order to further improve the accuracy and application range of pipeline network leakage detection, this paper simulates the complex water distribution system to set up an experimental platform. On this basis, a leakage detection method using improved wavelet denoising and short time Fourier transform is proposed, which can be used to monitor the leakage of the pipeline network in real time. This method is compared with the six methods of Empirical Mode Decomposition (EMD), wavelet decomposition, Gauss model, recurrence plot, Wigner-Ville distribution (WVD) and Wigner-Hough transformation (WHT). The results show that this method can accurately diagnose the leakage of the pipeline network. This method is verified by the field measured pressure signals of other literature. It is concluded that this method is effectual for the leakage detection of the actual complex pipeline network. In order to test the universality of this method, it is also used to analyze the pressure signal of the heating pipeline network and the gas pipeline network. The results show that the method proposed in this paper is also applicable to other complex pipeline networks.