Fig. 7: General framework of ST-GPINN for water quality prediction in WDSs. | npj Clean Water

Fig. 7: General framework of ST-GPINN for water quality prediction in WDSs.

From: ST-GPINN: a spatio-temporal graph physics-informed neural network for enhanced water quality prediction in water distribution systems

Fig. 7

Illustrates the sequential steps: a Schematic diagram for node concentration prediction, showing the workflow for spatial discretization, virtual node insertion, and iterative forward-backward optimization; b Details of the Encoder-Processor-Decoder architecture, capturing spatio-temporal dependencies and predicting node concentrations with physical consistency.

Back to article page