This elaborate proposes a methodology for implementing the Digital Twin (DT) of an industrial milling machine, to reproduce its energy behaviour and optimize its energy consumption (thus obtaining an Energy DT or EDT). The main gaps addressed include the lack of a comprehensive methodology for guiding DT development and deployment, as well as the need to expand the range of services an EDT can offer. Such gaps were addressed through the development of a structured deployment methodology which ensures a systematic evolution process, from the extraction of the raw data to the deployment of a fully operational EDT, as well as the integration of a Discrete Event Simulation (DES) for accurately simulating the Energy Consumption (EC) patterns. By also integrating a multi-objective scheduling optimization algorithm in the development of the EDT, the elaborate successfully addressed the Research Question: “How can a simulation-based Digital Twin be utilized to optimize energy consumption?”. In fact, the integration and collaboration between the DES and the optimization algorithm made the EDT capable of simulating very accurately EC patterns, and dynamically optimized the job scheduling. Such optimization was built upon the coexistence of two objectives to be optimized: total energy costs and total delays from missed deadlines of production. The Methodology was structured around four main phases: Data management & Trend identification (to polish and analyze the data), Key Parameters extraction (to extract from the data the specific parameters needed for the EDT development), EDT Deployment (to generate and benchmark solutions) and Validation (of the Methodology, through its own successful implementation). Performance benchmarking against heuristic methods was implemented, to validate the EDT superiority, corroborated with an ANOVA and Paired T-test implementation, revealing that the EDT developed managed to significantly save costs and reduce delays, highlighting the limitations of traditional scheduling methods and demonstrating the EDT’s capability to enhance energy-efficient production planning.
Questo elaborato propone una metodologia per l’implementazione del Digital Twin (DT) di una macchina fresatrice industriale, al fine di riprodurne il comportamento energetico e ottimizzarne il consumo energetico (ottenendo così un Energy DT o EDT). Le principali lacune affrontate includono l'assenza di una metodologia strutturata per guidare lo sviluppo e la distribuzione del DT, oltre alla necessità di ampliare la gamma di servizi che un EDT può offrire. Queste carenze sono state colmate attraverso lo sviluppo di una metodologia di implementazione strutturata, che garantisce un processo di evoluzione sistematico, dalla fase di estrazione dei dati grezzi fino alla distribuzione di un EDT completamente operativo, e mediante l’integrazione di un modello di Simulazione ad Eventi Discreti (SED) per simulare con precisione i modelli di Consumo Energetico (CE). Integrando inoltre un algoritmo di ottimizzazione multi-obiettivo per la schedulazione nella progettazione dell’EDT, l’elaborato ha risposto con successo alla domanda di ricerca: "Come può un Digital Twin basato su simulazione essere utilizzato per ottimizzare il consumo energetico?". L’integrazione e la collaborazione tra la SED e l’algoritmo di ottimizzazione hanno reso l’EDT capace di simulare con elevata precisione i modelli di consumo energetico e di ottimizzare dinamicamente la schedulazione dei lavori. L’ottimizzazione è stata realizzata considerando due obiettivi: la riduzione dei costi energetici totali e la minimizzazione dei ritardi nelle scadenze di produzione. La metodologia è stata strutturata in quattro fasi principali: Gestione dei dati e Identificazione dei trend (per pulire e analizzare i dati), Estrazione dei parametri chiave (per identificare i parametri essenziali allo sviluppo dell’EDT), Implementazione dell’EDT (per generare e confrontare le soluzioni) e Validazione (della metodologia, attraverso la sua effettiva implementazione).Il confronto delle prestazioni con metodi euristici è stato condotto per validare la superiorità dell’EDT, corroborato da un'analisi ANOVA e da un Paired T-test, dimostrando che l’EDT sviluppato ha consentito di ridurre significativamente i costi e i ritardi. Questi risultati evidenziano i limiti dei metodi di schedulazione tradizionali e dimostrano il potenziale dell’EDT nell’ottimizzazione della pianificazione produttiva in ottica di efficienza energetica.
Methodology for a digital twin-based energy consumption simulation and optimization
TATONI, LORENZO
2023/2024
Abstract
This elaborate proposes a methodology for implementing the Digital Twin (DT) of an industrial milling machine, to reproduce its energy behaviour and optimize its energy consumption (thus obtaining an Energy DT or EDT). The main gaps addressed include the lack of a comprehensive methodology for guiding DT development and deployment, as well as the need to expand the range of services an EDT can offer. Such gaps were addressed through the development of a structured deployment methodology which ensures a systematic evolution process, from the extraction of the raw data to the deployment of a fully operational EDT, as well as the integration of a Discrete Event Simulation (DES) for accurately simulating the Energy Consumption (EC) patterns. By also integrating a multi-objective scheduling optimization algorithm in the development of the EDT, the elaborate successfully addressed the Research Question: “How can a simulation-based Digital Twin be utilized to optimize energy consumption?”. In fact, the integration and collaboration between the DES and the optimization algorithm made the EDT capable of simulating very accurately EC patterns, and dynamically optimized the job scheduling. Such optimization was built upon the coexistence of two objectives to be optimized: total energy costs and total delays from missed deadlines of production. The Methodology was structured around four main phases: Data management & Trend identification (to polish and analyze the data), Key Parameters extraction (to extract from the data the specific parameters needed for the EDT development), EDT Deployment (to generate and benchmark solutions) and Validation (of the Methodology, through its own successful implementation). Performance benchmarking against heuristic methods was implemented, to validate the EDT superiority, corroborated with an ANOVA and Paired T-test implementation, revealing that the EDT developed managed to significantly save costs and reduce delays, highlighting the limitations of traditional scheduling methods and demonstrating the EDT’s capability to enhance energy-efficient production planning.File | Dimensione | Formato | |
---|---|---|---|
2025_04_Tatoni_Tesi.pdf
non accessibile
Dimensione
2.66 MB
Formato
Adobe PDF
|
2.66 MB | Adobe PDF | Visualizza/Apri |
2025_04_Tatoni_Abstract.pdf
non accessibile
Dimensione
352.78 kB
Formato
Adobe PDF
|
352.78 kB | Adobe PDF | Visualizza/Apri |
2025_04_Tatoni_Executive_Summary.pdf
non accessibile
Dimensione
1.11 MB
Formato
Adobe PDF
|
1.11 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/10589/236459