Using image pre-mapping for applications of monitoring electrical switchboards

The detection of possible faults in electrical installation by analyzing the electrical switchboard (ESB) is not currently performed efficiently since the problem is usually detected only when a qualified operator inspects the ESB, either by chance or only on the specified date of maintenance. Considering the frequent and current use of thermographic images for applications in electrical circuits and the need for more efficient and safe systems to detect problems in electrical installations, this work aimed to develop a method for detection and identification of faults in electrical installations through thermographic images of the ESB. This research proposes a method that can be considered as input to a monitoring network of an electrical system and aims at achieving simple computational processing and low cost applications as well as making it unnecessary for the operator to have any knowledge about thermal imaging, programming or thermographic camera use. Ten different scenarios of an ESB containing twenty-six circuit breakers were simulated. Using a supervised classifier, the images were classified according to color, defining the status of operation of each circuit-breaker. It was concluded that the developed classifier was efficient in detecting the status of operation of the ESB components, acting simply and reliably for the ten simulated scenarios.