A Probabilistic Digital Twin for Leak Localization in Water Distribution Networks Using Generative Deep Learning

. 2023 Jul 5;23(13):6179. doi: 10.3390/s23136179

Algorithm 1: Offline stage

Input:

N_{train}

, training hyperparameters, WAE-MMD architecture

1
Generate training samples, ${(x_{i}, c_{i})}_{i = 1}^{N_{train}}$ , by solving the forward problem (see Section 2.1);
2
Train the supervised WAE-MMD (see Section 2.3);

Output:

ϕ_{dec}