Nonlinear Mixed-Integer Heuristic Programming with Optimization Algorithm to Enhance the Water Distribution System

Low efficiency affects energy-demanding systems such as water distribution networks (WDNs). In these systems, the pressure typically is maintained below the switch using regulators to minimize water loss from leaks. Utilizing energy production equipment may be an effective way to reduce water waste, while also producing energy, although its viability depends on how much energy can be recovered. A water distribution system’s design is a combinatorial issue, which typically has a larger number of local optima. Consequently, hybrid metaheuristic and heuristic processes can explore the solution with less computational time requirements to gain deep understanding of the problem structure and specific characteristics of the problem. However, these outcomes have an enormous computational burden because of the relatively large number of hydraulic simulations. This investigated where pressure-reducing valves (PRVs) and pumps as turbines (PATs) should be placed in a network that distributes water. The study suggests a deterministic mathematical optimization technique for minimizing the price of WDNs utilizing recognized recognized pipe distances and a defined range of commercially obtainable sizes. A novel heuristic mixed-integer nonlinear programming technique (H-MINLP) with a beleaguered path search process is used to perform optimization. Based on the evaluation of the ideal trajectories in which water flows in a WDN. Two distinct subroutines work together to decrease the sizes of network pipes methodically and sequentially, and effectively exploit the search space. There are no parameters to adjust in the new method, and therefore it does not require a consequential purpose. In addition, a graph clustering technique is utilized to increase the heuristic approach’s performance by simplifying its convergence. Compared with other methods in the literature, this hybrid optimization ensures good solutions in terms of energy and water savings. According to the findings, compared with other research on the same network, the proposed optimization decreased leakage by 21%.