A Fast Electrical Distribution Fault Predictor using Knowledge Growing System (KGS)

Electrical energy is one of important needs in our everyday life from the household to the industries and increases day by day. Any disruption to the electrical distribution will disrupt its reliability that impacts to the electrical supply to various sectors. Such situation has often occurred at State Electrical Company (SEC) unit of Lambung Mangkurat in Borneo Island. Predicting the electrical distribution fault that causes the disruption can be carried out in advance to reduce the problem of electrical supply. For this purpose, we propose a new fast prediction technique called Knowledge Growing System (KGS) to predict the electrical distribution fault. KGS is an intelligent agent that can generate its own knowledge regarding a phenomenon it is observing and uses the generated knowledge to make predictions. By having knowledge about 11 electrical fault patterns at the electrical distribution site, KGS has able to predict that the most probable fault to occur is Fault Short-circuited Primary Coil Burned (FSPCB) with the probability of 0.2830. With a fast prediction, the SEC unit can develop a proper plan to cope with the fault and to recover the electrical supply quicker.