Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach

Tiku Tanyimboh, Calvin Siew

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

Abstract

In the context of water distribution system (WDS) design optimization, majority of the evolutionary algorithm models in the literature simulate solutions using demand driven analysis (DDA) based hydraulic solver and solutions that violate the nodal pressure constraints are penalized using penalty functions. The disadvantages of penalty functions are that they are case sensitive and require expertise in calibration with numerous time consuming trial runs. This paper investigates the application of a new penalty‐free multi‐objective evolutionary algorithm (PFMOEA) in the design optimization of WDS. Herein, the PFMOEA is extended to optimize problems involving pump operations, tank sizing and locations. The model succeeded in locating cheaper solutions than those previously reported in the literature that satisfy all the required constraints. Results are presented and compared to the best solution reported in the literature.
Original languageEnglish
Title of host publicationASCE/EWRI World Environmental & Water Resources Congress
DOIs
Publication statusPublished - May 2011
EventASCE/EWRI World Environmental & Water Resources Congress - Palm Springs, California, United States
Duration: 22 May 201126 May 2011

Conference

ConferenceASCE/EWRI World Environmental & Water Resources Congress
CountryUnited States
CityPalm Springs, California
Period22/05/1126/05/11

Fingerprint

Evolutionary algorithms
Water distribution systems
Systems analysis
Hydraulics
Pumps
Calibration
Design optimization

Keywords

  • Anytown network
  • evolutionary optimization approach
  • demand driven analysis

Cite this

Tanyimboh, T., & Siew, C. (2011). Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach. In ASCE/EWRI World Environmental & Water Resources Congress https://doi.org/10.1061/41173(414)20
Tanyimboh, Tiku ; Siew, Calvin. / Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach. ASCE/EWRI World Environmental & Water Resources Congress. 2011.
@inproceedings{b360202e1a284a4d9aca2f437a55368c,
title = "Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach",
abstract = "In the context of water distribution system (WDS) design optimization, majority of the evolutionary algorithm models in the literature simulate solutions using demand driven analysis (DDA) based hydraulic solver and solutions that violate the nodal pressure constraints are penalized using penalty functions. The disadvantages of penalty functions are that they are case sensitive and require expertise in calibration with numerous time consuming trial runs. This paper investigates the application of a new penalty‐free multi‐objective evolutionary algorithm (PFMOEA) in the design optimization of WDS. Herein, the PFMOEA is extended to optimize problems involving pump operations, tank sizing and locations. The model succeeded in locating cheaper solutions than those previously reported in the literature that satisfy all the required constraints. Results are presented and compared to the best solution reported in the literature.",
keywords = "Anytown network , evolutionary optimization approach , demand driven analysis",
author = "Tiku Tanyimboh and Calvin Siew",
year = "2011",
month = "5",
doi = "10.1061/41173(414)20",
language = "English",
booktitle = "ASCE/EWRI World Environmental & Water Resources Congress",

}

Tanyimboh, T & Siew, C 2011, Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach. in ASCE/EWRI World Environmental & Water Resources Congress. ASCE/EWRI World Environmental & Water Resources Congress, Palm Springs, California, United States, 22/05/11. https://doi.org/10.1061/41173(414)20

Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach. / Tanyimboh, Tiku; Siew, Calvin.

ASCE/EWRI World Environmental & Water Resources Congress. 2011.

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

TY - GEN

T1 - Design of the Anytown network using the penalty-free multi-objective evolutionary optimization approach

AU - Tanyimboh, Tiku

AU - Siew, Calvin

PY - 2011/5

Y1 - 2011/5

N2 - In the context of water distribution system (WDS) design optimization, majority of the evolutionary algorithm models in the literature simulate solutions using demand driven analysis (DDA) based hydraulic solver and solutions that violate the nodal pressure constraints are penalized using penalty functions. The disadvantages of penalty functions are that they are case sensitive and require expertise in calibration with numerous time consuming trial runs. This paper investigates the application of a new penalty‐free multi‐objective evolutionary algorithm (PFMOEA) in the design optimization of WDS. Herein, the PFMOEA is extended to optimize problems involving pump operations, tank sizing and locations. The model succeeded in locating cheaper solutions than those previously reported in the literature that satisfy all the required constraints. Results are presented and compared to the best solution reported in the literature.

AB - In the context of water distribution system (WDS) design optimization, majority of the evolutionary algorithm models in the literature simulate solutions using demand driven analysis (DDA) based hydraulic solver and solutions that violate the nodal pressure constraints are penalized using penalty functions. The disadvantages of penalty functions are that they are case sensitive and require expertise in calibration with numerous time consuming trial runs. This paper investigates the application of a new penalty‐free multi‐objective evolutionary algorithm (PFMOEA) in the design optimization of WDS. Herein, the PFMOEA is extended to optimize problems involving pump operations, tank sizing and locations. The model succeeded in locating cheaper solutions than those previously reported in the literature that satisfy all the required constraints. Results are presented and compared to the best solution reported in the literature.

KW - Anytown network

KW - evolutionary optimization approach

KW - demand driven analysis

UR - http://www.scopus.com/inward/record.url?scp=79960401504&partnerID=8YFLogxK

U2 - 10.1061/41173(414)20

DO - 10.1061/41173(414)20

M3 - Conference contribution book

BT - ASCE/EWRI World Environmental & Water Resources Congress

ER -