Enhanced Anomaly-Based Fault Detection System in Electrical Power Grids

Early and accurate fault detection in electrical power grids is a very essential research area because of its positive influence on network stability customer satisfaction. Although many techniques have been introduced during the past decade, existence an effective robust system still rare real-world applications. Moreover, one main challenges that delays progress this direction severe lack reliable data for validation. Therefore, paper proposes novel anomaly-based which consistent with concept faults grids. It benefits from two phases prior to training phase, namely, preprocessing pretraining. While phase executes all elementary operations raw data, pretraining selects optimal hyperparameters model using particle swarm optimization (PSO)-based algorithm. Furthermore, one-class support vector machines (OC-SVMs) principal component analysis (PCA) models are exploited validate proposed VSB dataset modern realistic dataset. Finally, results thoroughly discussed several quantitative statistical analyses. The experimental confirm effectiveness improving faults.