Real-time control (RTC) of wastewater systems has been a topic of research and application for over two decades. Attempts so far have mainly focused on one of the parts of the urban wastewater system: either the sewer system, or the treatment plant or the river. Approaches to integrate these subsystems and considering them jointly for control purposes have been pursued only recently. Control of the systems aims at pursuing one (or several concomitant) objectives, which are expressed, for example, in terms of overflow volumes, loads, effluent concentrations, receiving water quality or monetary costs, to name just a few. This paper provides a general and formal definition of the problem to define a real time control algorithm for a given urban wastewater system. A general mathematical optimization problem is formulated, which describes the task of finding an (in some sense) optimum control algorithm. Since this optimization problem is, in the general case, highly non-linear with only limited information available about the objective function itself, optimization methods appropriate for this type of problem are identified. Here, the similarity of the problem to find a control algorithm and of the parameter estimation problem common in mathematical modelling becomes apparent. Hence, methods (and problems encountered) in parameter estimation can be transferred to the problem of determining optimum RTC algorithms. This parallelism is outlined in the paper. As an application of the parameterisation and optimization of control strategies, integrated control of an urban wastewater system is discussed. Since the analysis of integrated control as just described poses certain requirements on a simulation engine, a novel modelling tool, called SYNOPSIS, is utilized here. This simulation tool, comprising of modules simulating water quantity and quality processes in all parts of the urban wastewater system, is embedded into a suite of optimization procedures. An integrated RTC algorithm for the urban wastewater system is formulated, the parameters of which are optimized using various global optimization routines. Comparison of their efficiency indicates good performance for the Controlled Random Search and for the genetic algorithms. The findings suggest that integrated control can indeed lead to an increase in performance of the urban wastewater system. These results appear to be encouraging and justify further work. Areas for further development are identified in the final section of the paper.
[1]
A. H. Lobbrecht,et al.
Dynamic Water-System Control
,
2020
.
[2]
Wolfgang Rauch,et al.
On the potential of genetic algorithms in urban drainage modeling
,
1999
.
[3]
W. Price.
Global optimization algorithms for a CAD workstation
,
1987
.
[4]
M. B. Beck.
Dynamic modelling and control applications in water quality maintenance
,
1976
.
[5]
M. B. Beck,et al.
Optimisation of control strategies for the urban wastewater system — An integrated approach
,
1999
.
[6]
W. Rauch,et al.
Real time control of wastewater systems
,
1996
.
[7]
W. J. Shuttleworth,et al.
Parameter estimation of a land surface scheme using multicriteria methods
,
1999
.
[8]
S. Sorooshian,et al.
Shuffled complex evolution approach for effective and efficient global minimization
,
1993
.
[9]
L. Lijklema.
Integrated Management of Urban Waters – A Preface
,
1993
.