Cyber-physical threat mitigation in wind energy systems: a novel secure architecture for industry 4.0 power grids

Abstract

In Industry 4.0, integrating Cyber-Physical Systems (CPS) within wind energy infrastructures introduces significant cyber-attack vulnerabilities. This paper presents the Hybrid Adaptive Threat Detection and Response System (HATDRS), a novel security architecture designed to enhance the resilience of wind energy systems against evolving cyber threats. The HATDRS model integrates a hybrid machine learning approach, combining supervised logistic regression with adaptive learning mechanisms, providing real-time threat detection and mitigation. This approach was chosen for its ability to integrate labelled data with real-time unsupervised feedback, providing dynamic and accurate threat detection in wind energy systems. The model was evaluated against traditional Intrusion Detection Systems (IDS) and Machine Learning-based Anomaly Detection Systems (ML-ADS) across key metrics, including accuracy, detection rate, false positive rate, response time, System Security Index (SSI), energy loss, and cost-efficiency. The results demonstrate that the HATDRS model outperforms its counterparts, achieving an accuracy of 95.4% and a detection rate of 97.2% while maintaining the lowest false positive rate (3.1%) and response time (500 ms). Additionally, the model achieved the highest SSI value of 88.7, significantly reducing energy loss to 1.5% and improving cost-efficiency to 0.528. These findings underscore the robustness and efficiency of the HATDRS model in mitigating cyber-physical threats in wind energy systems, offering a scalable and effective solution for securing renewable energy infrastructures. Future work will explore further optimization and real-world testing to validate the system’s scalability across diverse energy environments.