EUVIMEDSkopri EUVIMED
Open access · CC BY 4.0

Design and Model-in-the-Loop Validation of Data-Driven Predictive Electrical Load Control for a Laboratory-Scale Wind Turbine

Batın Demircan

Electronics · 2026-11-20 · 15 (22) · 5411

DOI: 10.3390/electronics15225411 ↗

At a glance

Journal article in Electronics, published online on 2026-11-20. Written by Batın Demircan. 25 references. The full text is freely available (open access, CC BY 4.0).

AstrattDettalji biblijografiċiCiteRelated studies

Abstract (original language)

This study develops a constrained, quasi-steady-state predictive electrical-load optimizer for maximum power point tracking in a laboratory-scale wind turbine. An artificial neural network power model, identified from experimental measurements, serves as the prediction engine: wind speed is the disturbance, the pitch angle is an offline configuration parameter, and the electrical-load command is the control input. A measurement matrix of nine wind speeds at the reference pitch (60°) and three speeds at each of three further angles (30°, 45°, 75°), with ten load levels, defined 180 conditions over two repetitions. One low-power condition was excluded, leaving 179 valid conditions (358 measurements). Repetition differences averaged 2.59% overall (3.53% reference, 1.65% further angles). A multi-layer perceptron using wind speed, load, and pitch angle attained R2 = 0.997 on the test set, and a SHAP analysis attributed 41% each to wind speed and pitch angle and 18% to load. Across the evaluated changing-pitch configurations, the optimizer achieved 96.1% mean power-capture efficiency, marginally exceeding the 2-percentage-point P&O baseline (95.8%). On fully held-out operating conditions, the retrained optimizer achieved 96.9% ± 1.3%, compared with 98.2% for model-free P&O. The identified first-order wind-side model reproduced an unseen multi-step wind profile with R2 = 0.978.

Cite

APA

Demircan, B. (2026). Design and Model-in-the-Loop Validation of Data-Driven Predictive Electrical Load Control for a Laboratory-Scale Wind Turbine. Electronics, 15(22), 5411. https://doi.org/10.3390/electronics15225411

Vancouver

Demircan B. Design and Model-in-the-Loop Validation of Data-Driven Predictive Electrical Load Control for a Laboratory-Scale Wind Turbine. Electronics. 2026;15(22):5411. doi:10.3390/electronics15225411

BibTeX
@article{Demircan2026,
  author = {Demircan, Batın},
  title = {Design and Model-in-the-Loop Validation of Data-Driven Predictive Electrical Load Control for a Laboratory-Scale Wind Turbine},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {22},
  pages = {5411},
  doi = {10.3390/electronics15225411}
}
Start for free →