Implementation of meta-heuristic and deep learning algorithms for power system cybersecurity

Power system cyber security is crucial due to their criticality. Cybersecurity is essential to protect vital infrastructure as power systems digitize. Meta-heuristic and deep learning techniques are used to improve power system cyber security in this paper. To evaluate their performance, the suggested approach is compared to traditional supervised machine learning algorithms including artificial neural networks (ANNs), convolutional neural networks (CNNs), and support vector machines (SVMs). The technique optimizes deep learning model hyper parameters and architectures to detect cyber risks. Cyberattacks on power systems can cause service outages and cascading failures with extensive social implications. Meta-heuristic and deep learning algorithms are integrated to improve power system cyber security in this study. Deep learning is good at pattern recognition and anomaly detection, while meta-heuristic algorithms optimize efficiently. A complete threat detection and mitigation strategy is proposed by merging these methodologies. The proposed methodology tests classic supervised machine learning algorithms such ANNs, CNNs, and SVMs. Simulations showed the algorithm worked better. It beat competition in accuracy, precision, recall, and F1-score.