The artificial intelligence (AI) is part of the modern Power Systems. It is used in protection and control of electrical lines and transformers with good results, in the future will be widely used for implementing the smart grid. Any research is getting closer to an autonomous system which can learn how to adapt in different situations during a normal or fault regime.
Optimal functioning of electrical distribution networks is to maintain the voltage into acceptable limits in order not to affect the insulation of the electric lines. During nominal regime the voltage variation is not so big due to the several voltage regulation methods, but during a fault regime, usually during earth-faults can occur over-voltages that can affect the lines’ insulation. That is why earth-faults must be identified and eliminated as quick as possible.
The article is researching the possibility to detect the earth-faults with fuzzy logic. The fuzzy logic is using artificial intelligence and allows the earth-faults detection even the input data is not so accurate because of the errors given by the instrument transformers. So, using several logic rules, the phase and the circuit with earth-fault can be successfully identified even the input data is not so precise.
The expert fuzzy systems can be successfully used in earth-fault detection and also in detecting any kind of fault with superior results to classical method because using the data with errors it can obtain good results using the logic small, normal, big for the currents and voltages in the system can take decisions similar to human operator.
The system using AI can cover wide areas of electrical networks, it can be used for disconnecting the other side of the line by communication between the protection relays.
