An Intelligent Lightning Prediction and Protection System Using ANN

Abstract

Lightning is a natural phenomenon that can cause damage to people and property. Usually, substation supervisors monitor weather conditions continuously, if there is a risk of damage to the equipment due to the inclement weather, then will decide to cut off the power supply manually by the recommendations of higher authority. But disconnecting the power supply during the heavy rain time creates unnecessary problems and accidents as well as leads to inefficient power management. Hence it is important to develop an intelligent system that can predict, alert and protect against lightning strikes. Here an attempt is made to develop an intelligent lightning prediction and protection system using ANN and IoT. The developed system consists of data collection, ANN training, lightning prediction and lightning protection phases. Here Artificial Neural Network (ANN) is specially used to improve the accuracy within a short period. ANN models are trained using input data, which will recognize the patterns and predict using testing data. The trained model is integrated into a lightning protection system i.e. IoT to take future actions. The hardware prototype is developed to disconnect the power supply automatically in the lightning-prone areas and also send the notification in the form of SMS to the operator in the respective zone.