Predict to prevent

This project intends to perform a temperature monitoring system of the components inside wind turbines distribution panel with the purpose of locating in advance the points likely to fail.

Nowadays this process lies in stopping the system, opening the distribution panel and taking a thermographic picture inside. Thanks to this operation, we can monitor the temperature of the distribution panel in real time without putting at risk the engineer and stopping the system. Besides the engineer’s display of the picture, the system has implemented a Cloud architecture that gets alerts and temperature data from it. Thus, predicting failures and fire risks is safer, more economical, reliable and traceable.

Technologies involved

For this project development we used microprocessors, photo cameras and low-cost thermographic cameras in order to offer interpretable pictures which can be accessed by WIFI network. A software containerization system allows to update the equipment remotely and storing it into the Cloud and databases. With this software, we received warnings which are sounded in the equipment either by maximum temperature or fast temperature rise. The communication among cameras and master equipment and between this master equipment and the Cloud system is realized by API REST.

Project Objectives

The wind turbines have a main problem which is the location. These wind turbines are usually in open field zones with a rough orographic accessibility, separated from urban areas and with significant distance among themselves. The control and monitoring condition of the wind turbines can be a difficult task when it’s necessary to check the condition of a specific parameter in all of the wind turbines belonging to an area. This project emerged for avoiding this kind of taks, speeding up the monitoring process and foreseeing the problems that an environment such as the wind turbine can generate.

Hitherto, an engineer goes periodically to the wind turbines one by one checking the distribution panels’ temperature with a thermal gun. The main target of this project is to automate this task, to do it in real time and to create ad-hoc alarms. And only when an alert is triggered, the engineer can check the pictures obtained in real time without having to switch off the equipment or preventing risks for the engineer.

Therefore, we seek to perform:

  • Distribution panels predictive maintenance.
  • Real time temperature monitoring.
  • Predictive failures based on temperature registration.
  • Automate generation of noise alerts due to high temperature changes.

Safe engineer monitoring during service.