Prediction and Feature-Level Analysis of Power Quality Indicators in Industrial Power Systems with Reactive Power Compensation Active and Inactive Using Ensemble Learning

Year 2026, Volume: 6 Issue: 1, 35 - 47, 27.02.2026

Faruk Kürker

https://doi.org/10.5152/tepes.2026.25035

https://izlik.org/JA98BX76GZ

Abstract

Reactive power compensation systems are widely used in industrial power grids for power factor improvement and reactive power optimization. However, the reliable estimation of power quality indicators under conditions where compensation is active or inactive is a critical engineering problem that has not been sufficiently addressed in the literature. This study proposes a prediction approach that provides high accuracy, interpretability, and noise robustness for both operating conditions. The input space consists of 19 physical parameters, including active (P), reactive (Q), and apparent (S) powers for each of the three phases, RMS current and voltage values per phase (I1rms-3rms and U1rms–3rms), interphase voltages (U12, U13, U23), and neutral current (Ineutral). The predicted targets are 11 fundamental power quality indicators: power factor (PF1–3, Pftot), displacement power factor (dPF1–3, dPFtot), and total harmonic distortion ratio per phase (ITHD1–3). The dataset was split into training and test subsets using a hold-out strategy, and independent bagging-based ensemble regression models were constructed for each target variable. Experimental results showed that the R² values ranged from 0.987 to 0.997 and the mean absolute percentage error (MAPE) values ranged from 0.68% to 3.57% for all targets. Normalized feature importance scores revealed the dominant role of RMS current components per phase. Model robustness was maintained with ΔR² < 0.004 under 5% Gaussian noise. The findings prove that the proposed method can predict power quality indicators with high reliability in both compensation cases and is suitable for predictive maintenance applications with real-time monitoring on an industrial scale.

Keywords

Ensemble regression , explainable artificial intelligence , noise robustness , power quality estimation , reactive power compensation systems

References

• 1. S. A. Giha Yidi, V. Sousa Santos, K. Berdugo Sarmiento, J. E. Candelo- Becerra, and J. de la Cruz, “Comparison of reactive power compensation methods in an industrial electrical system with power quality problems,” Electricity, vol. 5, no. 3, pp. 642–661, 2024.
• 2. R. S. Kuzmin, A. Zavalov, and S. V. Kuzmin, “Influence of reactive power compensation on power quality in grids up to 1000V,” in 2020 International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM), Sochi, Russia, 2020, pp. 1–5,
• 3. W. P. Guamán, G. N. Pesántez, M. A. Torres, S. Falcones, and J. Urquizo, “Optimal dynamic reactive power compensation in power systems: Case study of Ecuador–Perú interconnection,” Electr. Power Syst. Res., vol. 218, 109191, 2023.
• 4. S. Gu, F. Zhang, C. Pu, and X. Zhao, “Coordinated control of static reactive power generator and fixed capacitor bank under reactive power prediction,” in 2025 7th Asia Energy and Electr. Eng. Symp. (AEEES), 2025, pp. 1–6.
• 5. R. Fiorotti, H. R. O. Rocha, C. R. Coutinho, A. C. Rueda-Medina, A. F. Nardoto, and J. F. Fardin, “A novel strategy for simultaneous active/ reactive power design and management using artificial intelligence techniques,” Energy Convers. Manag., vol. 294, 117565, 2023.
• 6. J. Andruszkiewicz, J. Lorenc, and A. Weychan, “Determination of the optimal level of reactive power compensation that minimizes the costs of losses in distribution networks,” Energies, vol. 17, no. 1, p. 150, 2024.
• 7. A. Cataliotti, V. Cosentino, and S. Nuccio, “The measurement of reactive energy in polluted distribution power systems: An analysis of the performance of commercial static meters,” IEEE Trans. Power Deliv., vol. 23, no. 3, pp. 1296–1301, 2008.
• 8. K. Valiullin, and A. Kosenko, “Estimation of economic viability of reactive power compensation on gas production industry facilities, based on energy consumption data,” in 2024 Int. Ural Conf. Electr. Power Eng. (UralCon), 2024, pp. 97–102.
• 9. G. G. Tolun, Ö. C. Tolun, and K. Zor, “‘Advanced machine learning algorithms for reactive power forecasting in electric distribution systems,’ e-Prime – Adv. Electr. Eng., Electron,” Energy, vol. 13, 101019, 2025.
• 10. B. Karami, S. Galvani, M. Farhadi-Kangarlu, and A. Bagheri, “Optimal voltage set-point of automatic tap changer transformers and generators and reactive power compensation to increase power system predictability,” Electr. Power Syst. Res., vol. 230, 110244, 2024.
• 11. C. Ma, Q. Zhou, P. Zhang, and W. Zhang, “Intelligent algorithm-based reactive power compensation at distribution network for three-phase imbalance,” in 2012 Asia-Pacific Power and Energy Eng. Conf. (APPEEC), Shanghai, China, 2012, pp. 1–4.
• 12. L. Fang, A. Luo, X. Xu, and H. Fang, “Neural network and nonlinear prime-dual interior algorithm for optimization reactive power compensation,” in 2010 Int. Conf. Electr. and Control. Eng. (ICECE), Wuhan, China, 2010, pp. 3898–3901.
• 13. V. N. Costa, M. R. Monteiro, and A. C. Zambroni de Souza, “Particle swarm optimization applied to reactive power compensation,” in 2016 17th Int. Conf. Harmonics and Quality of Power (ICHQP), Belo Horizonte, Brazil, 2016, pp. 780–785.
• 14. H. Samet, and A. Mojallal, “Enhancement of electric arc furnace reactive power compensation using Grey–Markov prediction method,” IET Gener. Transm. Distrib., vol. 8, no. 9, pp. 1626–1636, 2014.
• 15. A. Rai, D. S. Babu, P. S. Venkataramu, and M. S. Nagaraja, “Artificial neural network application for prediction of reactive power compensation under line outage contingency,” in Int. Conf. Power, Energy and Control. (ICPEC), Dindigul, India, 2013, pp. 355–359.
• 16. Y. Lu, C. Lu, Y. Liu, M. Deng, Y. Chen, and R. Duan, “A multi-objective coordinating model for distribution network with EVs, energy storage, and reactive power compensation devices,” Energy Rep., vol. 12, pp. 2600–2610, 2024.
• 17. M. Karthikeyan, D. Manimegalai, and K. Rajagopal, “Enhancing voltage control and regulation in smart micro-grids through deep learningoptimized EV reactive power management,” Energy Rep., vol. 13, pp. 1095–1107, 2025.
• 18. M. L. T. Zulu, R. Sarma, and R. Tiako, “Modified fuzzy logic and artificial bee colony: An artificial intelligence approach to optimization and power quality improvement in an MPPT-based system,” Sci. Afr., vol. 28, e02690, 2025.
• 19. A. Bennagi, O. AlHousrya, D. T. Cotfas, and P. A. Cotfas, “Comprehensive study of the artificial intelligence applied in renewable energy,” Energy Strateg. Rev., vol. 54, 101446, 2024.
• 20. L. E. Aciu, and A. Dănilă, “An analysis of the power factor compensation within the power grid of an industrial plant by means of supervised learning methods,” in 2023 Int. Conf. Electromech. and Energy Syst. (SIELMEN), Craiova, Romania, 2023, pp. 1–6.
• 21. T. Hothorn, and B. Lausen, “Double-bagging: Combining classifiers by bootstrap aggregation,” Pattern Recognit., vol. 36, no. 6, pp. 1303–1309, 2003.
• 22. K. Özen, and D. Yıldırım, “Application of bagging in day-ahead electricity price forecasting and factor augmentation,” Energy Econ., vol. 103, 105573, 2021.
• 23. D. Chicco, M. J. WARRENS, and G. Jurman, “The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation,” PeerJ Comput. Sci., vol. 7, e623, 2021.