A new Digital Harmony Search algorithm for optimizing Pump Scheduling in Water Distribution Networks

Pumps in Water Distribution Networks (WDNs) adequately provide effective pressure where low elevation or high head losses are detected within the system. One of the most effective strategies to ensure economic sustainability is Pump Scheduling (PS), assuring the optimization of pump management and enabling significant energy cost saving. Meta-heuristic algorithms can be applied to Pump Scheduling, given their ability to provide reliable global solutions, further complemented by limited computational efforts. Nevertheless, they do not assure the achievement of the optimal global solution. In this paper, the Pump Scheduling optimization was solved by applying a modified Harmony Search (HS) algorithm, aimed at optimizing the energy consumption and the maintenance costs of pumps, by limiting the number of daily switches. The EPANET 2.0 solver was employed for hydraulic simulations and novel setting equations and penalties were included to limit the number of feasible solutions. The model was tested on two benchmark water networks, showing its capability to return near-optimal solutions within a short computational time.