Nicholas Nyaika, Cosmas Mwikirize (PhD), Andrew Katumba (PhD)3

DOI: https://doi.org/10.5281/zenodo.22772228

The increasing penetration of renewable energy sources introduces significant variability and uncertainty into grid-connected microgrids (MGs), affecting voltage stability, frequency regulation, power quality, fault ride-through capability, and operational resilience at the Point of Common Coupling (PCC). These challenges become more pronounced in networked MGs with changing electrical and communication topologies, heterogeneous operating conditions, communication constraints, and variable loads. This study proposes an adaptive, distributed control framework that integrates a Spatio-Temporal Graph Transformer (ST-GT) for learning evolving network relationships, cooperative Multi-Agent Reinforcement Learning (MARL) for decentralized control with physics-informed constraints, and differential-privacy-enhanced federated selective sharing for secure collaboration among MGs. The framework is evaluated on modified IEEE 33-bus and IEEE 69-bus networked MG systems using high-fidelity digital twins implemented in pandapower and pymgrid. Experiments consider renewable penetration of 50% to 80%, short-circuit ratios below 3, communication failures, forecast uncertainty, islanding, and fault ride-through disturbances. Compared with benchmark control approaches, the proposed framework reduces PCC voltage deviations and total harmonic distortion by 26% to 33%, achieves fault ride-through success rates above 96%, improves economic performance by 17% to 23%, and maintains robust operation under forecast errors exceeding 40%. These findings demonstrate that the framework provides a scalable, resilient, and privacy-preserving approach for coordinated control of networked grid-connected MGs.

Point of Common Coupling, Networked MGs, Spatio-Temporal Graph Transformer, Multi-Agent Reinforcement Learning.

This paper presented a novel framework termed ST-GTMARL-PIC-DP for uncertainty-resilient adaptive PCC control in networked grid-tied MGs. The proposed architecture integrates Spatio-Temporal Graph Transformers, Multi-Agent Reinforcement Learning, Physics-Informed Constraints, Federated Selective Sharing, and Differential Privacy within a unified learning framework. Comprehensive evaluations conducted using modified IEEE 33-bus and IEEE 69-bus systems demonstrated substantial improvements in voltage regulation, harmonic mitigation, fault ride-through capability, economic performance, and resilience against communication failures and renewable uncertainty. The proposed framework achieved reductions of approximately 26 to 33% These results demonstrate the feasibility of combining graph-transformer-based learning, physics-informed control, federated collaboration, and differential privacy to support the next generation of intelligent and resilient networked MGs. Overall, the proposed framework demonstrates superior scalability, resilience, privacy preservation, and operational efficiency compared with existing state-of-the-art approaches

[1] Carlo Fabrizio et al. “Power grid control with graphbased distributed reinforcement learning”. In: arXiv preprint arXiv:2509.02861 (2025).
[2] Renhai Feng et al. “Uniform physics informed neural network framework for microgrid and its application in voltage stability analysis”. In: IEEE Access 13 (2025), pp. 8110–8126.
[3] Wangyong Guo et al. “Learning-driven load frequency control for islanded microgrid using graph networksbased deep reinforcement learning”. In: Frontiers in Energy Research 12 (2024), p. 1517861.
[4] Mohamed Hassouna et al. “Graph reinforcement learning for power grids: A comprehensive survey”. In: Energy and AI (2026), p. 100671.
[5] Nima Khosravi et al. “Microgrid stability: A comprehensive review of challenges, trends, and emerging solutions”. In: International Journal of Electrical Power & Energy Systems 170 (2025), p. 110829.
[6] Hua Meng et al. “Enhanced reinforcement learningmodel predictive control for distributed energy systems: Overcoming local and global optimization limitations”. In: Building Simulation. Vol. 18. 3. Springer. 2025, pp. 547–567.
[7] Sayak Mukherjee et al. “Enhancing cyber resilience of networked microgrids using vertical federated reinforcement learning”. In: 2023 IEEE Power & Energy Society General Meeting (PESGM). IEEE. 2023, pp. 1–5.
[8] Kayode Ebenezer Ojo, Akshay Kumar Saha, and Viranjay Mohan Srivastava. “Review of advances in renewable energy-based microgrid systems: Control strategies, emerging trends, and future possibilities”. In: Energies 18.14 (2025), p. 3704.
[9] Neelofar Shaukat et al. “A Novel Physics-Constrained and Nature-Inspired Reinforcement Learning With Adaptive Model Predictive Control for Multi-Machine Renewable Rich Microgrids”. In: IET Smart Grid 9.1 (2026), e70075.
[10] Lizhi Wang et al. “Physics-informed, safety and stability certified neural control for uncertain networked microgrids”. In: IEEE Transactions on Smart Grid 15.1 (2023), pp. 1184–1187.

Nicholas Nyaika, Cosmas Mwikirize (PhD), Andrew Katumba (PhD)3 (2026). Spatio-Temporal Graph Transformer MARL for Resilient Point of Common Coupling Control in Networked Grid-Tied Micro-grids. International Journal of Computer Techniques, 13(5), 359–369. ISSN: 2394-2231. DOI: https://doi.org/10.5281/zenodo.22772228