Optimal Real-time Control of Sewer Networks

Recent years have seen a very marked increase in the desire to protect the environment from any and all malign influences. The maintenance or restoration of water quality is a vital part of that protection. A sine qua non of control system development for modern sewer networks is therefore the preservation of the water system around a network’s outflow(s). Several approaches have been proposed for the optimisation of sewage control and Optimal Real-time Control of Sewer Networks provides a comparative synthesis of a central sewer network flow control based on two of these: nonlinear-optimal and multivariable-feedback control.

In nonlinear optimal control, control and operational objectives are treated directly by the formulation of a nonlinear cost function minimized according to system constraints and the relevant state equation. The comparison presented uses the rolling horizon method for the real-time application of the optimal control algorithm with updated inflow predictions and updated initial conditions.

On the other hand, the linear multivariable feedback regulator – considered with and without feedforward terms to account for external inflows – is developed via a systematic linear-quadratic procedure including precise specifications on model structure, equations and the choice of nominal steady state and quadratic criterion.

The comprehensive testing and comparison of these protocols is undertaken on the basis of their respective control results for the real large-scale sewer network located around the river Obere Iller in Bavaria. The control strategies now implemented within this network are based on this study.

Starting at the selection of possible methods of control and moving to the actual implementation of those methods in a real sewer system, Optimal Real-time Control of Sewer Networks will be invaluable to control and civil engineers working in sewage flow and wastewater treatment and of great interest to academics wishing to see how their ideas on optimal control are likely to work out when practically applied.