In recent years, the development of new digital technologies has produced a remarkable impact on the industrial environment, leading to the new paradigm of the Industry 4.0. Among the many research areas opened by this modern scenario, simulation tools still cover a key role. Although the industrial landscape offers a wide availability of dedicated software tools, the complexity of real-world applications requires solution up to their de-mands, especially in the research area: flexibility in modeling, as well as ease in data ma-nipulation prove themselves as crucial requirements. It is under this perspective that this thesis work has been developed. The main goal pur-sued is to propose a modeling and simulation framework, based on Open-Source technol-ogies, able to overcome some of the restrictions that the traditional approach shows espe-cially when dealing with complex configurations, more suitable for research activities. In this view, a Modelica-based customizable simulation blocks library and a Python-based, data-driven and automated model generation module, have been designed to be mu-tually integrated, so as to feature straightforward implementation capabilities, adaptability to different configurations and elasticity. The final result is suitable for different expertise levels: an expert user has the opportunity to access the source code to customize it basing on necessity, but also non-skilled user can easily perform modeling and simulation and get access to the results. Eventually, the proposed implementation is compared to Simulink through a case study, to provide a benchmark and to prove that it could constitute a valuable alternative.
Negli ultimi anni, lo sviluppo delle nuove tecnologie digitali ha prodotto un notevole im-patto sul mondo industriale, e ha spinto verso il nuovo paradigma di Industria 4.0. Tra le numerose aree di ricerca aperte da questo moderno scenario, le tecniche di simula-zione ancora ricoprono un ruolo chiave. Sebbene in ambito industriale l’offerta di soft-ware dedicati sia ampia, la complessità delle applicazioni reali richiede soluzioni all’altezza, specialmente nel campo della ricerca: flessibilità nella modellazione e semplicità nella elaborazione dei dati si dimostrano requisiti fondamentali. È in questa prospettiva che il presente lavoro di tesi ha visto il suo sviluppo. Il principale obiettivo perseguito è quello di proporre una struttura di modellazione e simulazione, ba-sata su tecnologie “Open-Source”, in grado di superare alcune delle limitazioni che l’approccio tradizionale mostra, soprattutto nel confronto con configurazioni complesse, che sia più orientato ad attività di ricerca. In quest’ottica sono stati progettati una libreria customizzabile di blocchi per la simulazio-ne, basata sul linguaggio Modelica, e un modulo in linguaggio Python, basato sui dati, per la generazione automatica del modello, tra loro integrati in modo da garantire facile imple-mentazione, adattabilità a diverse configurazioni ed elasticità. Il risultato finale dimostra di potersi adattare a diversi livelli d’uso: un utente esperto ha la possibilità di accedere al codice sorgente per modificarlo in base alle necessità, mentre un utente senza competenze specifiche può facilmente completare le fasi di modellazione e simulazione ed accedere ai risultati. Infine, l’implementazione proposta è stata messa a confronto con Simulink attraverso un caso di studio, per fornire un benchmark e potersi dichiarare un’alternativa valevole.
Framework for open-source modeling and simulation of manufacturing systems
GAMBA, GIANLUCA
2019/2020
Abstract
In recent years, the development of new digital technologies has produced a remarkable impact on the industrial environment, leading to the new paradigm of the Industry 4.0. Among the many research areas opened by this modern scenario, simulation tools still cover a key role. Although the industrial landscape offers a wide availability of dedicated software tools, the complexity of real-world applications requires solution up to their de-mands, especially in the research area: flexibility in modeling, as well as ease in data ma-nipulation prove themselves as crucial requirements. It is under this perspective that this thesis work has been developed. The main goal pur-sued is to propose a modeling and simulation framework, based on Open-Source technol-ogies, able to overcome some of the restrictions that the traditional approach shows espe-cially when dealing with complex configurations, more suitable for research activities. In this view, a Modelica-based customizable simulation blocks library and a Python-based, data-driven and automated model generation module, have been designed to be mu-tually integrated, so as to feature straightforward implementation capabilities, adaptability to different configurations and elasticity. The final result is suitable for different expertise levels: an expert user has the opportunity to access the source code to customize it basing on necessity, but also non-skilled user can easily perform modeling and simulation and get access to the results. Eventually, the proposed implementation is compared to Simulink through a case study, to provide a benchmark and to prove that it could constitute a valuable alternative.| File | Dimensione | Formato | |
|---|---|---|---|
|
TESI_Gamba.pdf
non accessibile
Descrizione: tesi
Dimensione
9.72 MB
Formato
Adobe PDF
|
9.72 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/10589/165433