Deep Learning-Enabled Adaptive Control Of Power Converters For Next-Generation Renewable Energy Integration

Ayush Kumar; Baby  Khooshbu; Kamini  Kumari; Ajay Kumar; Ozair  Ahmad

doi:10.29284/g639pd34

Authors

Ayush Kumar Assistant Professor, Department of Electronics and Communication Engineering, Saharsa College of Engineering, Saharsa
Baby Khooshbu Assistant Professor, Department of Electronics and Communication Engineering, Saharsa College of Engineering, Saharsa.
Kamini Kumari Assistant Professor, Department of Electronics and Communication Engineering, Saharsa College of Engineering, Saharsa.
Ajay Kumar Assistant Professor, Department of Electrical Engineering, Saharsa College of Engineering, Saharsa.
Ozair Ahmad Assistant Professor, Department of Electronics and Communication Engineering, Government Engineering College, Kishanganj.

DOI:

https://doi.org/10.29284/g639pd34

Keywords:

Deep learning; adaptive control; power converters; renewable energy integration; reinforcement learning; LSTM; DDPG; maximum power point tracking.

Abstract

The accelerating penetration of solar photovoltaic (PV) and wind-based distributed generation has exposed the structural limitations of fixed-gain proportional–integral (PI) and classical linear controllers used in grid-interfacing power converters. This paper proposes a deep learning-enabled adaptive control framework that combines a Long Short-Term Memory (LSTM) feature extractor with a Deep Deterministic Policy Gradient (DDPG) reinforcement-learning agent to regulate DC-link voltage, track the maximum power point, and suppress harmonic distortion under rapidly varying irradiance, temperature, and load conditions. The proposed controller continuously learns an optimal duty-ratio/modulation policy from measured voltage, current, irradiance, and temperature states, eliminating the need for manual gain re-tuning. A MATLAB/Simulink-based hardware-representative model of a two-stage grid-connected converter was used to benchmark the proposed method against conventional PI, fuzzy-logic, and artificial neural network (ANN) controllers. Simulation results show that the proposed controller reduces settling time by up to 81.6%, overshoot by up to 87.3%, and total harmonic distortion (THD) by up to 70.7% relative to the conventional PI controller, while improving maximum power point tracking (MPPT) efficiency to 99.6%. These findings demonstrate that deep learning-enabled adaptive control offers a scalable and disturbance-resilient pathway for next-generation renewable energy integration.

References

1. Deghfel, N., Badoud, A. E., Al-Ahmadi, A. A., Bajaj, M., Zaitsev, I., & Ghoneim, S. S. M. (2024). Improving maximum power point tracking efficiency in solar photovoltaic systems using super-twisting algorithm and grey wolf optimizer. IET Renewable Power Generation, 18(15), 3329–3354. https://doi.org/10.1049/rpg2.13138

2. Hussaian Basha, H. (2024). A novel development of wide voltage supply DC–DC converter for fuel stack application with PSO-ANFIS MPPT controller. Scientific Reports, 14(1), Article 18826. https://doi.org/10.1038/s41598-024-69802-9

3. Idris, D. K. M., Birasa, R., Marțiș, C. S., & Philippe, D. R. (2024). Comparative analysis of classical and AI-based MPPT techniques in stand-alone PV systems: Perturb and observe, incremental conductance, and fuzzy logic. In Proceedings of the 2024 IEEE International Conference and Exposition on Electric and Power Engineering (EPEi) (pp. 139–144). IEEE. https://doi.org/10.1109/EPEi63510.2024.10758020

4. Pei, Y., Ye, K., Zhao, J., Yao, Y., Su, T., & Ding, F. (2024). Enhanced-visibility model-free deep reinforcement learning algorithm for voltage control in realistic distribution systems using smart inverters. Applied Energy, 359, Article 122716.

5. Rajamallaiah, A., Karri, S. P. K., Alghaythi, M. L., & Alshammari, M. S. (2024). Deep reinforcement learning based control of a grid connected inverter with LCL-filter for renewable solar applications. IEEE Access, 12, 22278–22295.

https://doi.org/10.1109/ACCESS.2024.3364058

6. Roy, B., Adhikari, S., Datta, S., Devi, K. J., Devi, A. D., & Ustun, T. S. (2024). Harnessing deep learning for enhanced MPPT in solar PV systems: An LSTM approach using real-world data. Electricity, 5(4), 843–860.

https://doi.org/10.3390/electricity5040042

7. Xing, Y., Zhang, G., Wang, B., et al. (2024). Modeling and deep reinforcement learning based control parameter tuning for voltage source converter in a renewable energy generation system. Journal of Electrical Engineering & Technology, 19, 947–959. https://doi.org/10.1007/s42835-024-01820-z

8. Younas, U., Kulaksiz, A. A., & Ali, Z. (2024). Deep learning stack LSTM based MPPT control of dual stage 100 kWp grid-tied solar PV system. IEEE Access, 12, 77555–77574.