Browsing by Author "Pecora, Pedro de Almeida"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Digital twin based condition monitoring system for a cold stamping machinePublication . Pecora, Pedro de Almeida; Leitão, Paulo; Agulhari, Cristiano MarcosThe fourth industrial revolution aims to connect and digitalize industrial assets. In an industrial environment, energy efficiency and machine health are important aspects of daily life. Energy monitoring systems, used alongside sensory data gathered from machines on the shop floor, can be powerful tools for monitoring the overall machine health. In this work, a data acquisition and monitoring system was implemented in a cold molding machine owned by Catraport, aiming to monitor electrical and vibrational data gathered during the molding process. The sensing equipment used communicates over Wi-Fi and, due to network problems and inconsistency, the factory’s computer network had to be modified, allowing for a better connection. The collected data was stored into InfluxDB, with timestamps for each measured input. With the collected data, five different dashboards were created using Grafana, one giving an overall view of the measured parameters, and the others containing more specific information of each parameter, namely current intensity, power, power factor and vibration. From the real-time data, out of control condition testing is carried out using Nelson Rules, imposing that whenever a parameter triggers one or more of the implemented rules, an alarm is shown on the dashboard and an email is sent to the maintenance technician. From the collected data a Machine Learning algorithm, named EECP-CBL, was implemented aiming to predict the next 5 minutes of current intensity, the forecasts also generate alerts for the maintenance team.