Model Reduction Using Neural Networks Applied to the Modeling of Integrated Urban Wastewater Systems

Simulation of the integrated urban wastewater system is a computationally-demanding task, and performing long-term simulations consequently takes a considerable time. Reduction of simulation times can be achieved by speeding up one or more of the submodels of the integrated system. In this paper the use of a fast neural network model instead of the mechanistic model of the wastewater treatment plant (WWTP) is proposed for this purpose. The neural network is trained on a sequence of treatment plant input/output data generated by the mechanistic model of the WWTP, i.e. it is a reduced model of the original WWTP model. As a result of model substitution a reduction of the simulation time by a factor of 23 was achieved. The results presented in this paper show that the errors introduced by the WWTP model substitution are of an acceptable level, confirming the practical usefulness of the proposed method.