Optimal Control of a Sewer Network

Vasiliev, Iulian; Luca, Laurențiu; Vilanova, Ramón; Caraman, Sergiu

doi:10.1109/icstcc55426.2022.9931810

Cited by 2 publications

(1 citation statement)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the papers (Joseph‐Duran et al., 2014, 2015; Ocampo‐Martinez et al., 2007; Sun et al., 2021; Troutman et al., 2020) propose predictive algorithms in order to minimize in real time the amount of polluted water. A series of papers (Minzu & Serbencu, 2016; Minzu et al., 2015; Ochoa et al., 2019; Rathnayake & Faisal Anwar, 2019; Safavi & Geranmehr, 2017; Vasiliev et al., 2022b, 2022c) presents different optimization algorithms such as Particle Swarm Optimization (PSO), genetic algorithms, FminSearch optimization etc. aiming to simultaneously minimize two objectives: the pollutant loads in the wastewater discharged from the SN directly into the environment and the cost of the WWT and of the energy consumption.…”

Section: Introductionmentioning

confidence: 99%

Reducing the Environmental Impact of Sewer Network Overflows Using Model Predictive Control Strategy

Vasiliev,

Luca,

Barbu

et al. 2024

Water Resources Research

View full text Add to dashboard Cite

This paper proposes a method for reducing the environmental impact of sewer network (SN) overflows. The main objective of the paper is to minimize the wastewater quantity and the pollutant loads that overflow from the SN. The proposed algorithm to achieve this goal is Model Predictive Control using Particle Swarm Optimization as optimization method. It was tested in simulation using a simplified model of the network based on Benchmark Simulation Modelsewer as prediction model, and a forecasted influent. Three cases have been considered: (a) the fitness function is defined as the global yearly overflow volume calculated using equal weights for each tank; (b) the fitness function uses different weights for each tank depending on the medium loads and (c) integrating a penalty term related to the system state at the end of the prediction horizon in the previous fitness function. The simplified model determined a significant reduction of the integration time minimizing the optimization time.

show abstract