Iot-Based Intelligent Fault Analysis of Distribution Transformers with Aquila Optimization and Secure Transformer Attention Network

Abstract

Distribution Transformers (DT) is used to activate dependably near a stable supply of electricity in smart grids, where unplanned malfunction results in expensive outages and tackle damage. In this paper proposed online fault analysis system is recognized on the Internet of Things (IoT) and tolerates for real-time transformer state monitoring as well as analysis. The pre-processing is smeared to the obtained sensor data, which contains data cleaning to eliminate noise and irregularities as well as Z-score normalization to standardise variances across various measurement scales. The Aquila Optimization Algorithm (AOA), it efficiently finds the most discriminative attributes associated with fault patterns, is used to perform feature selection to improve diagnostic accuracy and lower processing cost. And a proposed advanced Deep Learning (DL) architecture designed to perform accurate and interpretable fault analysis using Secure Transformer Attention Network (STAN). Experimental assessments validate that proposed technique develops fault localisation accuracy as well as recall of 96.7%, precision of 96.8% by python software, increases strength to data variants. Strength of the paper is by using the optimization technique it improves the performance.