Digital and bioinspired approaches to designing sustainable educational spaces: modeling, optimization, and adaptive systems

Aida Anafina, Alla Aleksandrovna Kornilova

Sustainable Engineering and Innovation, 2026

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.37868/sei.v8i2.id1020
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

It’s an ideal milieu for this study which proposes a mathematically rigors, digitally bio-inspired framework for sustainable educational space design. It incorporates parametric Building Information Modeling (BIM), building energy modeling, computational fluid dynamics (CFD), daylight and acoustic analytics, surrogate response modeling, bioinspired multi-objective optimization, entropy weighting combined with TOPSIS ranking, alpha-cut uncertainty propagation, interval type-2 fuzzy adaptive control. The application of the framework to a prototypical 4860 m² learning center described by fifteen decision variables (including envelope geometry, fenestration with shading and thermal mass) is presented. Herein, the numerical design database for the high reducing iron concentration is composed of 420 Latin hypercube design points along with 800 Monte Carlo uncertainty realizations. The optimization problem formulation seeks to minimize the building's energy use intensity, discomfort hours, operational carbon, acoustic penalty and daylight deficiency while maximizing spatial flexibility following graph-based metrics. Results compare Genetic Algorithm, Self-Adaptive Particle Swarm Optimization and multi-objective Ant Colony Optimization via hypervolume (IHV), GD+, spacing, run-time, and a composite pedagogic sustainability index adjusted for reliability. The chosen particle-swarm method reduces energy use intensity from 154.2 kWh/m²·yr to 93.7 kWh/m²·yr, operational carbon from 127.8 kgCO2e/m²·yr to 75.4 kgCO2e/m²·yr, discomfort hours from 418 h/year to143 h/year and reverberation time from 0.92 seconds to0.59s while enhancing useful daylight illuminance (UDI) value from 54% to 81%. Compared with fixed rule-based operation, the interval type-2 fuzzy controller provides an additional 17.6% in daily energy savings. Architecturally, the framework is interpreted through learning-space typology, zoning, circulation, daylight access, acoustic privacy, and user experience, so the mathematical layer supports spatial design decisions rather than replacing architectural judgment.