Modelamiento conceptual y metodología para el desarrollo de un gemelo digital orientado a la ciberseguridad del tacómetro óptico en aerogeneradores de ELECAUSTRO S.A.

Abstract

This research paper details the development of a cybersecurity-oriented Digital Twin for the optical tachometer in PMSG wind turbines at the Huascachaca Wind Farm, operated by ELECAUSTRO S.A., with the objective of mitigating critical infrastructure vulnerabilities against threats such as False Data Injection (FDI) attacks that can destabilize the grid. To achieve this, the methodology establishes an architecture structured into five layers (Physical, Communication, Processing, Simulation, and Visualization) that achieves a successful convergence between operational technology (OT) and information technology (IT). On one hand, it employs MATLAB/Simulink to simulate the turbine's electromechanical behavior with high fidelity by integrating FOC and VOC controls, which ensures extremely high-quality power delivery with a total harmonic distortion (THD) of just 0.04%. On the other hand, communication is managed via low-latency TCP/IP sockets to a Python server, where the Isolation Forest Artificial Intelligence algorithm analyzes telemetry and detects any anomalous behavior in real time (in under 50 milliseconds) without the need for prior supervision, immediately displaying alerts on a visual and interactive dashboard. Ultimately, this unified model functions as a secure testing environment (testbed) that allows for the evaluation of cyber threats without compromising real equipment, thereby strengthening cyber resilience and guaranteeing the continuity of the electrical supply to society.