Ecology-Aware Search in a Three-Species Chemostat: Production and Stability
Systemic Analytics · 26.09.2026 · 4 (3) · 235-250
DOI: 10.31181/sa43202693 ↗Kurz gefasst
Fachartikel in der Zeitschrift Systemic Analytics, online erschienen am 26.09.2026. Verfasst von Seyyed Esmaeil Najafi und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
A deterministic three-species chemostat benchmark compares five production and ecological screening objectives using six continuous variables. ECO-MAOEA combines fixed-total inoculum normalization, a contracting feasibility tolerance, and reference-point selection. Across the reported 3,000-evaluation, ten-seed comparison, its feasibility rate exceeds the other evolutionary implementations (0.468 ± 0.032 versus 0.307 for NSGA-II and 0.161 for MOEA/D), while its three-objective hypervolume is lower (0.102 ± 0.016 versus 0.286 ± 0.013 for MOEA/D). Removing normalization also changes total inoculum, preventing an isolated repair-effect claim. Under substrate inhibition, ECO-MAOEA has the highest descriptive feasibility rate, 0.440 ± 0.047, while coverage remains lowest. Selected fronts show production–spectral-margin trade-offs; feasibility does not certify stability. The findings concern the recorded surrogate and implementations. Replication requires original code and per-run outputs.
Zitieren
APA
Najafi, S. E., Karim, R. G., & Muniz, R. D. F. (2026). Ecology-Aware Search in a Three-Species Chemostat: Production and Stability. Systemic Analytics, 4(3), 235-250. https://doi.org/10.31181/sa43202693
Vancouver
Najafi SE, Karim RG, Muniz RDF. Ecology-Aware Search in a Three-Species Chemostat: Production and Stability. Systemic Analytics. 2026;4(3):235-250. doi:10.31181/sa43202693
BibTeX
@article{Najafi2026,
author = {Najafi, Seyyed Esmaeil and Karim, Ramin Goudarzi and Muniz, Rita de Fátima},
title = {Ecology-Aware Search in a Three-Species Chemostat: Production and Stability},
journal = {Systemic Analytics},
year = {2026},
volume = {4},
number = {3},
pages = {235--250},
doi = {10.31181/sa43202693}
}
EUVIMED