Prior Assumptions for Leak Localisation in Water Distribution Networks with Uncertainties

Abstract Hydraulic model-based leak (burst) localisation in water distribution networks is a challenging problem due to limited number of hydraulic measurements, wide range properties, and model data uncertainties. In this study, prior assumptions are investigated improve the presence For example, $$\ell _2$$ ℓ 2 -regularisation relies on assumption that Euclidean norm coefficient vector should be minimised. This approach compared with method based sensitivity matrix, which assumes existence only single leak. The results show while matrix often yields better location estimate scenarios, successfully identifies search area for pinpointing accurate location. Furthermore, it shown additional error introduced by quadratic approximation Hazen-Williams formula solution negligible given uncertainties resistance coefficients operational network models.