Training Doubly Fed Induction Generator Performance Parameters Using Variational Quantum Regressor and Classifier Models

Renewable energy systems, particularly wind energy conversion systems, are increasingly adopting advanced data-driven techniques for performance modelling and control. Among wind generators, the doubly fed induction generator (DFIG) is widely used due to its high efficiency, variable-speed operation, and four-quadrant capability. In this work, Variational Quantum Machine Learning (QML) models are employed to analyse key DFIG performance parameters. The rotor current and reactive power are selected as target outputs and modelled using a Variational Quantum Classifier (VQC) and Variational Quantum Regressor (VQR), respectively. Four critical DFIG design and operating features are encoded into quantum feature space using Qiskit-based feature maps. The quantum models are trained on a dataset of 1000 specially developed DFIG designs spanning a power range of 1000 kW to 2100 kW. Model performance is evaluated in terms of prediction accuracy and convergence behaviour, highlighting the feasibility and potential advantages of variational quantum algorithms for power system component modelling in the near-term quantum computing era.