The integration of volatile renewable energy sources into modern power grids presents significant challenges for maintaining stability and efficiency. This paper presents a climate-aware multi-agent reinforcement learning (MARL) system for smart grid energy management, specifically designed for the diverse regional grids of India. The proposed framework integrates an LSTM-based forecasting module for predicting short-term renewable generation with a custom MARL environment where agents learn collaborative policies for grid optimization. We evaluate the system using a realistic synthetic dataset generated from the energy profiles of four key Indian regions: Tamil Nadu, Odisha, Rajasthan, and Bihar. Experimental results demonstrate that the MARL framework successfully learns distinct, region-specific operating policies, achieving stable global grid balance through coordinated thermal generation management. The integration of climate forecasts enables proactive rather than reactive control strategies, significantly enhancing grid resilience. This work provides a foundational framework for developing adaptive, resilient energy management systems for renewable-integrated power grids in developing economies.

This paper presented a climate-aware multi-agent reinforcement learning framework for smart grid energy management, specifically designed for India’s diverse regional grids. The framework integrates an LSTM-based forecasting module for renewable generation prediction with a custom MARL environment where regional agents learn collaborative control policies. Evaluation on a realistic synthetic dataset representing Tamil Nadu, Odisha, Rajasthan, and Bihar demonstrated that the system successfully learns distinct, region-appropriate operating policies while maintaining global grid balance. The LSTM forecasts proved critical, enabling proactive rather than reactive control and improving performance by 40-57% across key metrics. This work provides a foundational framework for developing adaptive, resilient energy management systems capable of addressing the complex challenges of renewable-integrated power grids in developing economies.

[1] L. Xie, P. M. S. Carvalho, L. A. F. M. Ferreira, J. Liu, B. H. Krogh, N. Popli, and M. D. Ilić, “Wind integration in power systems: Operational challenges and possible solutions,” Proceedings of the IEEE, vol. 99, no. 1, pp. 214–232, Jan. 2011. [2] F. Capitanescu, J. L. Martinez Ramos, and P. Panciatici, “State-of-the-art, challenges, and future trends in security constrained optimal power flow,” Electric Power Systems Research, vol. 81, no. 8, pp. 1731–1741, Aug. 2011. [3] M. Morari and J. H. Lee, “Model predictive control: past, present and future,” Computers & Chemical Engineering, vol. 23, no. 4–5, pp. 667–682, May 1999. [4] L. Busoniu, R. Babuska, and B. De Schutter, “A comprehensive survey of multiagent reinforcement learning,” IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), vol. 38, no. 2, pp. 156–172, Mar. 2008. [5] Indian Smart Grid Forum, “National Smart Grid Mission,” Ministry of Power, Government of India. [Online]. Available: https://www.nsgm.gov.in/. [Accessed: 15-Mar-2026]. [6] A. J. Conejo, E. Castillo, R. Minguez, and R. Garcia-Bertrand, Decomposition techniques in mathematical programming: engineering and science applications. New York, NY, USA: Springer, 2006. [7] D. Q. Mayne, J. B. Rawlings, C. V. Rao, and P. O. M. Scokaert, “Constrained model predictive control: Stability and optimality,” Automatica, vol. 36, no. 6, pp. 789–814, June 2000. [8] D. Ernst, M. Glavic, and L. Wehenkel, “Power systems stability control: reinforcement learning framework,” IEEE Transactions on Power Systems, vol. 19, no. 4, pp. 1871–1880, Nov. 2004. [9] R. Lu, S. H. Hong, and M. Yu, “Demand response for home energy management using reinforcement learning and artificial neural network,” IEEE Transactions on Smart Grid, vol. 10, no. 6, pp. 6629–6639, Nov. 2019. [10] M. T. J. Spaan, “Partially observable Markov decision processes,” in Reinforcement Learning: State-of-the-Art, M. Wiering and M. van Otterlo, Eds. Berlin, Germany: Springer, 2012, pp. 387–414. [11] J. Li, D. Shi, and T. Yang, “Multi-agent reinforcement learning for economic dispatch with practical constraints,” IEEE Transactions on Power Systems, vol. 36, no. 4, pp. 3267–3278, July 2021. [12] S. Chen, Y. Liu, and M. Zhang, “A multi-agent reinforcement learning approach for voltage control in distribution networks,” IEEE Transactions on Power Systems, vol. 37, no. 2, pp. 1404–1415, Mar. 2022. [13] R. Lowe, Y. Wu, A. Tamar, J. Harb, P. Abbeel, and I. Mordatch, “Multi-agent actor-critic for mixed cooperative-competitive environments,” in Proc. Advances in Neural Information Processing Systems, Long Beach, CA, USA, 2017, pp. 6379–6390. [14] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, Nov. 1997. [15] Y. Wang, D. Gan, N. Zhang, and C. Kang, “A review of deep learning for renewable power forecasting,” CSEE Journal of Power and Energy Systems, vol. 7, no. 5, pp. 945–957, Sept. 2021. [16] Z. Wang, X. Liu, and Y. Huang, “Short-term photovoltaic power forecasting based on LSTM network with attention mechanism,” IEEE Access, vol. 9, pp. 31924–31933, Feb. 2021. [17] H. Liu, X. Mi, and Y. Li, “Wind speed forecasting method based on deep learning strategy using empirical wavelet transform, long short term memory and neural network,” Energy Conversion and Management, vol. 156, pp. 498–514, Jan. 2018. [18] X. Zhang, Y. Bao, and D. Zhao, “Reinforcement learning for smart grid control: A review,” IEEE Access, vol. 9, pp. 51731–51749, Apr. 2021. [19] S. K. Soonee, S. R. Narasimhan, and V. K. Agrawal, “Indian power system: Challenges and opportunities,” IEEE Power and Energy Magazine, vol. 15, no. 4, pp. 34–42, July 2017. [20] P. Paliwal, N. P. Patidar, and R. K. Nema, “Planning of grid integrated distributed generators: A review of technology, objectives and techniques,” Renewable and Sustainable Energy Reviews, vol. 40, pp. 584–597, Dec. 2014. [21] J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proximal policy optimization algorithms,” arXiv preprint arXiv:1707.06347, 2017.