Predicting Cyber Attacks on Integrated Energy and Gasoline Resources

Connecting gas and electricity networks has recently been proposed as a viable solution to manage the unpredictability of renewable energy sources and increase the versatility of power operations. When critical operational communication and control inputs from both systems must be shared, the likelihood of cybercrime increases. This study introduces a novel approach, the Energy and Gasoline Integration for Cyber-Attacks (EGICA) model, which addresses cybersecurity in interconnected gas and electricity networks—a solution to manage renewable energy unpredictability and enhance operational flexibility. The model illustrates how undetected cyberattacks can compromise network operations by using power-to-gas (PtG) and gas-to-generation (GtG) methods. Two innovative detection techniques are proposed for identifying hazardous false data in PtG/GtG schedulers. The first employs fully convolutional and wavelet transformations to detect attacks in facilities planning data. In contrast, the second utilizes a hybrid neuron with unsupervised learning to detect attacks on scheduler outputs without labelled training data. These methods support a robust cyber-attack scheduling framework within integrated gas and electricity networks. Simulation tests on a bus energy system connected to a utility grid validate the model’s effectiveness using historical data for evaluation. Results show that EGICA achieves high accuracy (86.3%), precision (89.1%), recall (87.2%), F-measure (90.2%), and efficiency (78.7%) compared to other methods, confirming its reliability in detecting cyber threats in complex energy networks.