Optimising pump system with constant and variable speed pumps: case study

Optimising pump scheduling is a complex problem, which involves a large space search, continuous and discrete variables, physical and operational constraints and also multi-objectives. In this paper, multi-objective evolutionary algorithms (MOEAs) combined with a repair mechanism are used to solve the optimal operation problem within water supply system. In this work two objectives are minimised: operation cost (energy cost + treatment cost) and maintenance cost, while one objective is maximised: service level of hydraulic. Decision variables are the settings of the pumps and speed ratio of variable-speed pumps at a time step of the total operational time horizon. A mixed coding methodology and a new crossover operator are developed according to the characteristics of decision variables. Three well-known MOEAS (NSGA-II, epsilon-MOEA and SPEA2) are implemented and compared. Practical application of this method shows that it can make efficient decision to support the operators.