doi:10.3808/jei.200600064
Copyright © 2022 ISEIS. All rights reserved

Benchmark Tests of Evolutionary Algorithms: Mathematic Evaluation and Application to Water Distribution Systems

     B. S. Jung1*, B. W. Karnev2 and M. F. Lambert3

  1. MWH Soft, 300 N. Lake Avenue Pasadena, CA 91101, USA
  2. Department of Civil Engineering, Univ. of Toronto, Toronto, ON M5S 1A4, Canada
  3. Department of Civil and Environmental Engineering, University of Adelaide, Adelaide SA 5005, Australia

     *Corresponding author. Email: BongSeog.Jung@mwhsoft.com

Abstract


Evolutionary Algorithms (EAs) are a set of probabilistic optimization algorithms based on an analogy between natural biological systems and engineered systems. In this paper, the computational performance a set of specific EAs (specifically, the Genetic Algorithm, Evolutionary Programming, Particle Swarm Optimization, Ant Colony Optimization and Shuffled Complex Evolution Algorithm) are compared using a set of four mathematical test objective functions. In addition, a hybridization of EAs with other local search methods is introduced to improve or fine-tune the performance the primary EA. As a case study, the EAs are applied to a calibration problem for a water distribution system and ably show their robust and global convergence characteristics.

Keywords: Benchmark, calibration, evolutionary algorithms, hybridization


Full Text:

PDF

Supplementary Files:

Refbacks

  • There are currently no refbacks.