Multiobjective optimisation of urban wastewater systems using ParEGO: a comparison with NSGA II

Abstract

Commercial and research-based simulation models are now available to represent the performance and control of the sewer network, wastewater treatment plant and receiving water as a whole. To improve overall system performance, these models can be combined with optimisation methods to derive optimal control strategies. The popular evolutionary algorithms (EAs) have been proven to be a powerful method in developing optimal control strategies; however, the high computational requirements of these methods impose a limit on their application due to the complexity of the system. This paper explores the potential of a surrogate based multi-objective optimisation method, ParEGO, for real time control of urban wastewater systems. An existing integrated model is used to evaluate the multiple objectives. This method is compared with NSGA II by using two performance indicators: the hypervolume indicator and the additive binary ε ε-indicator. Comparative results show that ParEGO is an efficient and effective method in deriving optimal control strategies for the multiple objective control problems. It is suggested that ParEGO can greatly improve the computational efficiency, particularly for complex systems.