Author(s): Chi Zhang, Aaron Zecchin, Jinzhe Gong, Martin Lambert, Angus Simpson

Linked Author(s): Chi Zhang

Keywords: Inverse Transient Analysis (ITA), parameter estimation, noise, pipeline condition, wavespeed.

Abstract: Hydraulic transient based methods have been widely studied for anomaly detection and pipeline condition assessment in water distribution systems. Among existing hydraulic transient based methods, Inverse Transient Analysis (ITA) has the potential to achieve cost-effectiveness and spatially continuous assessment of transmission pipelines. In ITA, system parameters are estimated by minimizing the difference between a measured and simulated pressure trace. Measured pressure traces in the field are inevitably contaminated by hydraulic noise. A coarse grid is typically used in ITA to limit the number of decision variables and the model complexity. Research into ITA parameter estimation accuracy, subject to hydraulic noise and damaged section whose length is shorter than the resolution used in ITA, is undertaken in this paper. The investigations are based on a simulated reservoir-pipe-valve model, where a randomly fluctuating head at the reservoir is used to simulate different levels of hydraulic noise. The wavespeeds of the pipes are estimated using the noisy measured pressure traces by the ITA approach. The wavespeeds estimated by ITA approach were compared to the true wavespeeds of the model. ITA was shown to be robust for the numerical case study in this paper

DOI:

Year: 2017