This thesis project proposes an advanced methodological framework for investigating building envelope and HVAC design variables in nearly Zero Energy Buildings (nZEBs), demonstrating the critical importance of deploying adequate computational tools from the earliest stages of the design process. The distinct advantage of this methodology lies in its multi-objective optimization approach, which enables the simultaneous analysis of performance metrics rarely examined together in literature: the minimization of energy consumption, alongside the maximization of both thermal comfort (PMV) and daylight availability (UDI). The parametric model, developed within the Rhinoceros and Grasshopper environments, was applied to the LabZEB experimental laboratory at the ITC-CNR in San Giuliano Milanese and validated across three distinct Italian macroclimates, represented by Milan (Continental climate), Rome (Mediterranean climate), and Aosta (Alpine climate). The findings reveal that the optimal building configuration varies significantly depending on the season and the specific climatic context, with the sole exception of the Window-to-Wall Ratio (WWR), whose optimal value consistently converges at 30% across all investigated locations. Finally, two variants of the NSGA-II genetic algorithm, coupled with Radial Basis Function (RBF) surrogate models, were implemented and compared against a discrete Grid Search baseline. The results exhibit negligible deviations between the methods, highlighting only marginal precision gains for the continuous evolutionary workflow in extreme climates. This ultimately demonstrates that the adoption of highly complex computational models is not always justified by a significant increase in overall accuracy.
Il presente lavoro di tesi si propone come uno strumento metodologico avanzato per lo studio di variabili progettuali e impiantistiche per edifici a energia quasi zero (nZEB), dimostrando l'importanza dell'utilizzo di strumenti di calcolo adeguati sin dalle prime fasi progettuali. Il vantaggio di questa metodologia risiede nell'approccio di ottimizzazione multi-obiettivo, consentendo l'analisi di obiettivi raramente trattati simultaneamente in letteratura: la minimizzazione dei consumi energetici, la massimizzazione del comfort termo-igrometrico (PMV) e del comfort visivo (UDI). Il modello parametrico, sviluppato negli ambienti Rhinoceros e Grasshopper, è stato applicato al laboratorio sperimentale LabZEB dell'ITC-CNR presso la sede di S. Giuliano Milanese, e validato su tre macroclimi italiani, rappresentati dalle località di Milano (clima continentale), Roma (clima mediterraneo) e Aosta (clima alpino). La scelta dell’assetto ideale varia significativamente in base alla stagione e al contesto climatico, ad eccezione del rapporto vetrato, il cui valore ottimo si attesta al 30% per tutte le località indagate. Infine, due varianti dell'algoritmo genetico NSGA-II accoppiate a modelli surrogati basati sulle Radial Basis Functions (RBF), vengono implementate e confrontate con la baseline a griglia discreta (Grid Search). I risultati mostrano scostamenti trascurabili tra i due metodi, evidenziando solo marginali benefici a favore dell'approccio continuo nei climi estremi, a dimostrazione che l'impiego di modelli più complessi non è sempre giustificato da un significativo aumento dell'accuratezza nei risultati.
Quadro metodologico per l'ottimizzazione multi-obiettivo di edifici a energia quasi zero (nZEB): analisi comparativa tra solutori a griglia discreta e algoritmi evolutivi
SCELLATO, SANIA
2025/2026
Abstract
This thesis project proposes an advanced methodological framework for investigating building envelope and HVAC design variables in nearly Zero Energy Buildings (nZEBs), demonstrating the critical importance of deploying adequate computational tools from the earliest stages of the design process. The distinct advantage of this methodology lies in its multi-objective optimization approach, which enables the simultaneous analysis of performance metrics rarely examined together in literature: the minimization of energy consumption, alongside the maximization of both thermal comfort (PMV) and daylight availability (UDI). The parametric model, developed within the Rhinoceros and Grasshopper environments, was applied to the LabZEB experimental laboratory at the ITC-CNR in San Giuliano Milanese and validated across three distinct Italian macroclimates, represented by Milan (Continental climate), Rome (Mediterranean climate), and Aosta (Alpine climate). The findings reveal that the optimal building configuration varies significantly depending on the season and the specific climatic context, with the sole exception of the Window-to-Wall Ratio (WWR), whose optimal value consistently converges at 30% across all investigated locations. Finally, two variants of the NSGA-II genetic algorithm, coupled with Radial Basis Function (RBF) surrogate models, were implemented and compared against a discrete Grid Search baseline. The results exhibit negligible deviations between the methods, highlighting only marginal precision gains for the continuous evolutionary workflow in extreme climates. This ultimately demonstrates that the adoption of highly complex computational models is not always justified by a significant increase in overall accuracy.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_07_Scellato_Tesi.pdf
accessibile in internet per tutti a partire dal 29/06/2027
Descrizione: Tesi
Dimensione
16.49 MB
Formato
Adobe PDF
|
16.49 MB | Adobe PDF | Visualizza/Apri |
|
2026_07_Scellato_Executive Summary.pdf
accessibile in internet per tutti a partire dal 29/06/2027
Descrizione: Executive Summary
Dimensione
2.25 MB
Formato
Adobe PDF
|
2.25 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/260858