Abstract
Power Station (PS) monitoring systems are becoming critical, ensuring electrical safety through early warning, and in the event of a PS fault, the power supply is quickly disconnected. Traditional technologies are based on relays and don't have a way to capture and store user data when there is a problem. The proposed framework is designed with the goal of providing smart environments for protecting electrical types of equipment. This paper proposes an Internet of Things (IoT)-based Smart Framework (SF) for monitoring the Power Devices (PD) which are being used in power substations. A Real-Time Monitoring (RTM) system is proposed, and it uses a state-of-the-art smart IoT-based System on Chip (SoC) sensors, a Hybrid Prediction Model (HPM), and it is being used in Big Data Processing (BDP). The Cloud Server (CS) processes the data and does the data analytics by comparing it with the historical data already stored in the CS. No-Structural Query Language Mongo Data Base (MDB) is used to store Sensor Data (SD) from the PSs. The proposed HPM combines the Density-Based Spatial Clustering of Applications with Noise (DBSCAN)-algorithm for Outlier Detection (OD) and the Random Forest (RF) classification algorithm for removing the outlier SD and providing Fault Detection (FD) when the PD isn't working. The suggested work is assessed and tested under various fault circumstances that happened in PSs. The simulation outcome proves that the proposed model is effective in monitoring the smooth functioning of the PS. Also, the suggested HPM has a higher Fault Prediction (FP) accuracy. This means that faults can be found earlier, early warning signals can be sent, and the power supply can be turned off quickly to ensure electrical safety. A powerful RTM and event warning system can also be built into the system before faults happen.