and analytical methods for smart grids.
Analyse IR still images of selected grid elements (protection switching equipment), detecting temperature changes due to current flow changes (different load) or due to fault conditions. During fault conditions protection element opens, plasma may be formed, high temperature differential is observed. Protection element can leak gas (SF6) when it is exhibiting performance degradation (ageing). Detecting such conditions on a series of numerous images, with automatic pattern recognition (protection element) and IR image analysis done only on the protection element temperature differential. Images are acquired from different sites (20 kV/0.4 kV), 24/7. Embedded platform runs Linux on ARM or is a virtual machine (KVM, VMWare). What features / clusters are detectable? Fault / error detection regarding the temperature changes and identification of the relevant clusters / selected grid elements only. How can these be used to minimize data traffic between IR camera (local embedded computer) and central server?
Algorithms, methods, alarm notification, sw program.