Leakage detection and calibration of pipe networks by the inverse transient analysis modified by Gaussian functions for leakage simulation

Abstract

The Inverse Transient Analysis (ITA) is a well known method for leakage detection and calibration of pipe networks. To reduce the problem of dimensionality as well as to allocate candidate leakages everywhere in the network and to handle the simulation and measurement uncertainties, it is assumed that a leakage has a quasi-normal distribution around its true location. Accordingly, a Gaussian function is introduced to simulate each candidate leakage so that the Gaussian function parameters are the ITA decision variables. To manage the ITA process and decrease unnecessary computations, a conceptual step-by-step algorithm is introduced through which the number of candidate leakages is gradually increased until the convergence criteria are met. Solving a benchmark example reveals that the modifications play a significant role in increasing both accuracy and efficiency of the ITA. As compared to the traditional ITA, the new version was successful in solving the example with about 30 times less computation.