| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.47 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A indústria 4.0 é um conceito recente e cada vez mais aplicado na otimização de processos
de fabrico. Dentre as tecnologias mais utilizadas, o digital twin apresenta uma forma de se
monitorizar e controlar remotamente processos industriais utilizando ferramentas de modelagem
3D e comunicação de dados, criando-se modelos virtuais do processo industrial.
Estes são então alimentados com os dados reais, caraterizando assim um gêmeo digital do
sistema real. Este trabalho apresenta o desenvolvimento do digital twin de um sistema
de manufatura flexível composto por quatro células de usinagem e um robô manipulador.
Para o desenvolvimento do sistema foram inicialmente desenvolvidos modelos 3D de todo
o sistema de manufatura utilizando-se ferramentas de desenho CAD 3D, então exportados
para o ambiente de simulação V-REP. Os modelos foram alimentados com os dados
coletados em tempo real do controlador do sistema através do desenvolvimento de uma
aplicação de aquisição de dados em JAVA, e o desenvolvimento de scripts no ambiente de
simulação, dinamizando o sistema virtual. Foi desenvolvido um sistema de monitorização
de dados e detecção de falhas para auxiliar o usuário em tomadas de decisão, utilizando
a plataforma Node-Red. Após a conclusão do trabalho, foi possível a visualização remota
do sistema, monitorização de dados em tempo real e alarmes de falhas caso haja algum
comportamento inesperado do sistema. O uso de sistemas computacionais que possibilitem
menor latência é encorajado, de forma a aplicar a tecnologia em aplicações que
necessitam de maior precisão, tal como uso de sistemas de machine learning para otimizar
algoritmos de detecção de falhas.
Industry 4.0 is a recent and increasingly applied concept into optimization of manufacturing processes. Among the most used technologies, the digital twin presents a way to remotely monitor and control industrial processes, using 3D modeling and data communication tools to create virtual models of industrial process. These are then fed with real data, therefore characterizing a digital twin of the real system. This work presents the development of the digital twin of a flexible manufacturing system composed of four machining cells and a manipulator robot. For the development of the system, 3D models of the entire manufacturing system were created using 3D CAD drawing tools, then exported to the V-REP simulation environment. The models were fed with data collected in real time from the system’s controller through the development of a data acquisition application in JAVA. The development of scripts in the simulation environment enabled the dynamizing of the virtual system. A data monitoring and fault detection system was developed for assisting the users in decision making, using Node-Red. After the work’s conclusion, the remote system visualization, real-time data monitoring and activation of alarms were made possible, in the event of an unexpected system behavior. The use of computer systems that allow for less latency is encouraged, in order to apply the technology to applications that need higher accuracy, such as using machine learning techniques to optimize the fault detection algorithms.
Industry 4.0 is a recent and increasingly applied concept into optimization of manufacturing processes. Among the most used technologies, the digital twin presents a way to remotely monitor and control industrial processes, using 3D modeling and data communication tools to create virtual models of industrial process. These are then fed with real data, therefore characterizing a digital twin of the real system. This work presents the development of the digital twin of a flexible manufacturing system composed of four machining cells and a manipulator robot. For the development of the system, 3D models of the entire manufacturing system were created using 3D CAD drawing tools, then exported to the V-REP simulation environment. The models were fed with data collected in real time from the system’s controller through the development of a data acquisition application in JAVA. The development of scripts in the simulation environment enabled the dynamizing of the virtual system. A data monitoring and fault detection system was developed for assisting the users in decision making, using Node-Red. After the work’s conclusion, the remote system visualization, real-time data monitoring and activation of alarms were made possible, in the event of an unexpected system behavior. The use of computer systems that allow for less latency is encouraged, in order to apply the technology to applications that need higher accuracy, such as using machine learning techniques to optimize the fault detection algorithms.
Description
Mestrado de dupla diplomação com a UTFPR - Universidade Tecnológica Federal do Paraná
Keywords
Sistema flexível de fabrico Indústria 4.0