Skip to content

Deep Learning Meta-Models for Water Systems Optimisation

The research utilised an adapted multi-objective optimisation algorithm called NSGA-II, adding onto this algorithm machine-learning based meta-models in a manner based on Learning Evolution Models for Multiple Objectives (LEMMO).

This created the LEMMO-ANN algorithm and optimised drainage networks in terms of reduction of flood risk (measured economically as expected annual damage), and reduced of cost of necessary modifications to the drainage network being optimised.

quote icon

by Dr William Sayers​ and Dr Thomas Win​

link icon

This research forms part of the Applied Business and Technology research priority area.