Detection of Cyber Attacks in Smart Grid Using Optimization and Deep Learning Techniques

Smart Grids are complex and interconnected systems that are vulnerable to cyber-attacks, which can result in severe consequences such as equipment damage, data theft, and power outages. Therefore, developing effective cyber-attack detection techniques for smart grids is crucial to ensure their security and reliability. This research propose a novelty for cyber-attack identification in smart grids using genetic optimization algorithm and deep feed forward neural network techniques. The proposed approach consists of three main components, data preprocessing, genetic optimization, and deep feed forward neural networks technique. The data preprocessing implicates cleansing the raw input data, removing noise, and extracting relevant features from the input data obtained from various sources in smart grids. The genetic algorithm is habituated to enhance the weight function and bias signals of the deep feedforward neural network to reduce the fault rate between the estimated output and the actual output. The deep feedforward neural network is responsible for detecting cyber-attacks based on preprocessed input data. To scrutinize the effectiveness of the proposed technique, conduct experiments on a publicly available dataset of smart grids. The observational outcome exhibits that the proposed model compares the existing techniques, obtaining a higher discovery depreciate and lower false alarm depreciate.