\n \n \n
\n
\n\n \n \n \n \n \n Generative AI and LLM applications in renewable energy and smart grids: a systematic review for the sustainable energy transition.\n \n \n \n\n\n \n Cali, U.; Halden, U.; Andoni, M.; Catak, F. O.; Chen, S.; Couraud, B.; Kantar, E.; Knapper, S.; Kucukdemiral, i.; Kusetogullari, H.; Kuzlu, M.; Mousavi, Y.; Norbu, S.; Ustun, T. S.; and Flynn, D.\n\n\n \n\n\n\n
Artificial Intelligence Review. March 2026.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@article{95f3ca0ecac246d7b73c0679fdb3b14a,\n\ttitle = "Generative AI and LLM applications in renewable energy and smart grids: a systematic review for the sustainable energy transition",\n\tabstract = "The global energy transition toward decarbonization and digitalization requires advanced methods to manage decentralized, data-intensive cyber-physical energy systems. This systematic review analyzes 106 research studies on Generative Artificial Intelligence (GenAI) and Large Language Models (LLMs) in renewable energy and smart grids, organized into seven application clusters covering forecasting, system design, operation, reliability, data and cybersecurity, and energy markets. The review situates these applications within a Cyber-Physical-Social Systems (CPSS) framework. Results show that GANs dominate current applications (47.2\\%), followed by LLMs (10.4\\%) and VAEs (9.4\\%), with growing adoption of diffusion and score-based models (7.5\\% each). Selected studies report improved probabilistic forecasting and uncertainty calibration using diffusion and score-based approaches, subject to dataset and evaluation setup. GenAI supports system planning through synthetic scenario generation, enhances operational decision support and demand response coordination, and contributes to reliability, cybersecurity, and market analysis. LLMs primarily function as language-driven decision support and knowledge integration components across multiple application domains. Despite computational and data-related constraints, GenAI represents an important enabler of the sustainable digital transition by supporting resilience, adaptability, and governance in renewable energy systems.",\n\tkeywords = "Cyber-Physical-Social Systems, Energy Digitalization, Energy Transition, Generative Artificial Intelligence, Large Language Models, Sustainable Energy Systems, Smart Grids, Renewable Energy",\n\tauthor = "Umit Cali and Ugur Halden and Merlinda Andoni and Catak, Ferhat Ozgur and Si Chen and Benoit Couraud and Emre Kantar and Samuel Knapper and ibrahim Kucukdemiral and Huseyin Kusetogullari and Murat Kuzlu and Yashar Mousavi and Sonam Norbu and Ustun, Taha Selim and David Flynn",\n\tyear = "2026",\n\tmonth = mar,\n\tday = "20",\n\tlanguage = "English",\n\tjournal = "Artificial Intelligence Review",\n\tissn = "0269-2821",\n\tpublisher = "Springer Netherlands",\n}\n\n\n
\n\n\n
\n The global energy transition toward decarbonization and digitalization requires advanced methods to manage decentralized, data-intensive cyber-physical energy systems. This systematic review analyzes 106 research studies on Generative Artificial Intelligence (GenAI) and Large Language Models (LLMs) in renewable energy and smart grids, organized into seven application clusters covering forecasting, system design, operation, reliability, data and cybersecurity, and energy markets. The review situates these applications within a Cyber-Physical-Social Systems (CPSS) framework. Results show that GANs dominate current applications (47.2%), followed by LLMs (10.4%) and VAEs (9.4%), with growing adoption of diffusion and score-based models (7.5% each). Selected studies report improved probabilistic forecasting and uncertainty calibration using diffusion and score-based approaches, subject to dataset and evaluation setup. GenAI supports system planning through synthetic scenario generation, enhances operational decision support and demand response coordination, and contributes to reliability, cybersecurity, and market analysis. LLMs primarily function as language-driven decision support and knowledge integration components across multiple application domains. Despite computational and data-related constraints, GenAI represents an important enabler of the sustainable digital transition by supporting resilience, adaptability, and governance in renewable energy systems.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Energy-aware hierarchical fractional-order terminal sliding mode control with hybrid learning observer for enhanced gait comfort in active prosthetics.\n \n \n \n\n\n \n Mousavi, A.; Mousavi, R.; Mousavi, Y.; Tavasoli, M.; Shahnazinia, S.; Küçükdemiral, I. B.; Fekih, A.; and Çalı, Ü.\n\n\n \n\n\n\n
Biomedical Signal Processing and Control, 113: 108933. 2026.\n
\n\n
\n\n
\n\n
\n\n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{MOUSAVI2026108933,\n\ttitle = {Energy-aware hierarchical fractional-order terminal sliding mode control with hybrid learning observer for enhanced gait comfort in active prosthetics},\n\tjournal = {Biomedical Signal Processing and Control},\n\tvolume = {113},\n\tpages = {108933},\n\tyear = {2026},\n\tissn = {1746-8094},\n\tdoi = {https://doi.org/10.1016/j.bspc.2025.108933},\n\tauthor = {Arash Mousavi and Rashin Mousavi and Yashar Mousavi and Mahsa Tavasoli and Sara Shahnazinia and Ibrahim Beklan Küçükdemiral and Afef Fekih and Ümit Çalı},\n\tkeywords = {J},\n}\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n FracTrace: Long-Horizon Credit Assignment in Deep Reinforcement Learning via Fractional Eligibility Traces.\n \n \n \n\n\n \n Mousavi, Y.; Mousavi, A.; Mousavi, R.; Tavasoli, M.; Küçükdemiral, I. B.; Fekih, A.; and Çalı, Ü.\n\n\n \n\n\n\n
IEEE Transactions on Neural Networks and Learning Systems. 2026.\n
Under review\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{Mousavi2026FracTrace,\n\tauthor = {Yashar Mousavi and Arash Mousavi and Rashin Mousavi and Mahsa Tavasoli and Ibrahim Beklan Küçükdemiral and Afef Fekih and Ümit Çalı},\n\ttitle = {FracTrace: Long-Horizon Credit Assignment in Deep Reinforcement Learning via Fractional Eligibility Traces},\n\tjournal = {IEEE Transactions on Neural Networks and Learning Systems},\n\tnote = {Under review},\n\tyear = {2026}\n}\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Safety-critical data-enabled predictive control for wheeled mobile robot.\n \n \n \n \n\n\n \n Erüst, A. C.; Taşcıkaraoğlu, F. Y.; and Küçükdemiral, İ. B.\n\n\n \n\n\n\n
ISA Transactions. 2026.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{CANERUST2026,\n\ttitle = {Safety-critical data-enabled predictive control for wheeled mobile robot},\n\tjournal = {ISA Transactions},\n\tyear = {2026},\n\tissn = {0019-0578},\n\tdoi = {https://doi.org/10.1016/j.isatra.2026.01.035},\n\turl = {https://www.sciencedirect.com/science/article/pii/S0019057826000492},\n\tauthor = {Ali Can Erüst and Fatma Yıldız Taşcıkaraoğlu and İbrahim Beklan Küçükdemiral},\n\tkeywords = {Data-enabled predictive control, Discrete time control barrier functions, Adaptive regularization, Wheeled mobile robot},\n\tkeywords = {J},\n}\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n A Safety-Preserving Event-Triggered Control Framework for Learning-Based Model Predictive Control.\n \n \n \n\n\n \n Erüst, A. C.; Taşcıkaraoğlu, F. Y.; and Küçükdemiral, İ. B.\n\n\n \n\n\n\n
Under Review. 2026.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{CANERUST2026RAS,\n\ttitle = {A Safety-Preserving Event-Triggered Control Framework for Learning-Based Model Predictive Control},\n\tjournal = {Under Review},\n\tyear = {2026},\n\tauthor = {Ali Can Erüst and Fatma Yıldız Taşcıkaraoğlu and İbrahim Beklan Küçükdemiral},\n\tkeywords = {J},\n}\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Model Predictive Control: From Foundations to Advanced Topics.\n \n \n \n \n\n\n \n Küçükdemiral, İ.\n\n\n \n\n\n\n Glasgow, UK, 2026.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@book{Kucukdemiral2026,\n\ttitle = "Model Predictive Control: From Foundations to Advanced Topics",\n\tauthor = "Küçükdemiral, İbrahim",\n\tyear = 2026,\n\taddress = "Glasgow, UK",\n\turl = "https://a.co/d/07UyNaBY",\n}\n\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n SAFE-WECS-LLM: A Multi-Layered Safety Validation Framework for Large Language Models in Critical Wind Energy Applications.\n \n \n \n\n\n \n Mousavi, Y.; Tavasoli, M.; Kucukdemiral, I. B.; Arab, A.; Fekih, A.; and Çalı, Ü.\n\n\n \n\n\n\n
Artificial Intelligence Review. 2026.\n
Under Review\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{ArtificialIntellegence2025_yashar,\nauthor = {Yashar Mousavi and Mahsa Tavasoli and Ibrahim Beklan Kucukdemiral and Aliasghar Arab and Afef Fekih and Ümit Çalı},\ntitle = {SAFE-WECS-LLM: A Multi-Layered Safety Validation Framework for Large Language Models in Critical Wind Energy Applications},\njournal = {Artificial Intelligence Review},\nyear = {2026},\nnote = {Under Review},\nkeywords = {J},\n}\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Dynamic Event-Triggered Robust Model Predictive Control for Quadrotor Trajectory Tracking.\n \n \n \n \n\n\n \n Erüst, A. C.; Taşcıkaraoğlu, F. Y.; and Küçükdemiral, İ. B.\n\n\n \n\n\n\n
International Journal of Robust and Nonlinear Control, 36(6): 3676-3688. 2026.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@article{https://doi.org/10.1002/rnc.70351,\n\tauthor = {Erüst, Ali Can and Taşcıkaraoğlu, Fatma Yıldız and Küçükdemiral, İbrahim Beklan},\n\ttitle = {Dynamic Event-Triggered Robust Model Predictive Control for Quadrotor Trajectory Tracking},\n\tjournal = {International Journal of Robust and Nonlinear Control},\n\tvolume = {36},\n\tnumber = {6},\n\tpages = {3676-3688},\n\tkeywords = {dynamic event-triggered control, quadrotor, robust model predictive control, trajectory tracking},\n\tdoi = {https://doi.org/10.1002/rnc.70351},\n\turl = {https://onlinelibrary.wiley.com/doi/abs/10.1002/rnc.70351},\n\tabstract = {ABSTRACT This paper addresses the trajectory tracking problem for a full-state quadrotor subject to physical model constraints and unknown external disturbances. A robust tube-based model predictive control (MPC) approach is successfully applied to the system, which is subject to bounded disturbances and hard constraints. In the literature, to reduce the computational complexity of standard time-triggered (ET) MPC without sacrificing performance, ET-MPC has been proposed, solving the optimal control problem only when an event is triggered. In this study, a dynamic threshold set is determined based on the worst-case disturbance effect and the deviation between the actual and predicted states of the quadrotor. Additionally, the discrete-time model of the quadrotor is extended with integral action, enabling the quadrotor to track the reference trajectory without error. To demonstrate the effectiveness of the proposed method, simulation results for time-triggered tube MPC and tube-based dynamic ET-MPC are compared. The proposed method proves its computational efficiency without compromising trajectory tracking performance. Moreover, the dynamic event trigger reduces the computational load 65\\%–85\\%, with an acceptable level of control performance degradation.},\n\tyear = {2026}\n}\n\n\n\n\n
\n\n\n
\n ABSTRACT This paper addresses the trajectory tracking problem for a full-state quadrotor subject to physical model constraints and unknown external disturbances. A robust tube-based model predictive control (MPC) approach is successfully applied to the system, which is subject to bounded disturbances and hard constraints. In the literature, to reduce the computational complexity of standard time-triggered (ET) MPC without sacrificing performance, ET-MPC has been proposed, solving the optimal control problem only when an event is triggered. In this study, a dynamic threshold set is determined based on the worst-case disturbance effect and the deviation between the actual and predicted states of the quadrotor. Additionally, the discrete-time model of the quadrotor is extended with integral action, enabling the quadrotor to track the reference trajectory without error. To demonstrate the effectiveness of the proposed method, simulation results for time-triggered tube MPC and tube-based dynamic ET-MPC are compared. The proposed method proves its computational efficiency without compromising trajectory tracking performance. Moreover, the dynamic event trigger reduces the computational load 65%–85%, with an acceptable level of control performance degradation.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Using time delayed disturbance compensation for sliding mode control.\n \n \n \n \n\n\n \n Han, X.; Küçükdemiral, İ.; Fridman, E.; Hakvoort, W. B J; Jamieson, J.; and Suphi Erden, M.\n\n\n \n\n\n\n
Journal of the Franklin Institute, 363(1): 108259. 2026.\n
\n\n
\n\n
\n\n
\n\n \n \n
Paper\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{HAN2026108259,\n\ttitle = {Using time delayed disturbance compensation for sliding mode control},\n\tjournal = {Journal of the Franklin Institute},\n\tvolume = {363},\n\tnumber = {1},\n\tpages = {108259},\n\tyear = {2026},\n\tissn = {0016-0032},\n\tdoi = {https://doi.org/10.1016/j.jfranklin.2025.108259},\n\turl = {https://www.sciencedirect.com/science/article/pii/S0016003225007513},\n\tauthor = {Xiaoran Han and İbrahim Küçükdemiral and Emilia Fridman and Wouter B J Hakvoort and James Jamieson and Mustafa {Suphi Erden}},\n\tabstract = {Under the framework of using Time Delay Control (TDC) for disturbance compensation in Sliding Mode Control (SMC), we address two practical problems. The first problem involves mitigating chattering in SMC caused by input delay, while the second problem focuses on designing TDC under conditions of limited measurements. Our research demonstrates that incorporating TDC as a phase lead compensator in the first problem can effectively accommodate larger input delays. To reduce the chattering, we propose a switching gain design, consisting of reference signals rather than measured states, resulting in an ultimately bounded solution. In the second problem, when acceleration measurements are unavailable, we provide stability conditions under which TDC can be designed with acceleration construction using delayed velocity signals. By constructing a modified sliding surface that incorporates the integral error remainder associated with the acceleration construction, our approach ensures switching gain are kept at minimal, with robust disturbance compensation at higher frequencies. We perform a simulation investigation of an autonomous underwater vehicle under input delay and disturbance to demonstrate the efficiency of the approach.},\n\tkeywords = {J},\n}\n\n\n
\n\n\n
\n Under the framework of using Time Delay Control (TDC) for disturbance compensation in Sliding Mode Control (SMC), we address two practical problems. The first problem involves mitigating chattering in SMC caused by input delay, while the second problem focuses on designing TDC under conditions of limited measurements. Our research demonstrates that incorporating TDC as a phase lead compensator in the first problem can effectively accommodate larger input delays. To reduce the chattering, we propose a switching gain design, consisting of reference signals rather than measured states, resulting in an ultimately bounded solution. In the second problem, when acceleration measurements are unavailable, we provide stability conditions under which TDC can be designed with acceleration construction using delayed velocity signals. By constructing a modified sliding surface that incorporates the integral error remainder associated with the acceleration construction, our approach ensures switching gain are kept at minimal, with robust disturbance compensation at higher frequencies. We perform a simulation investigation of an autonomous underwater vehicle under input delay and disturbance to demonstrate the efficiency of the approach.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n QRL-AFOFA: Q-Learning Enhanced Self-Adaptive Fractional Order Firefly Algorithm for Large-Scale and Dynamic Multiobjective Optimization Problems.\n \n \n \n\n\n \n Mousavi, Y.; Akbari, P.; Mousavi, R.; Mousavi, A.; Küçükdemiral, I.; Fekih, A.; and Çalı, Ü.\n\n\n \n\n\n\n
Artificial Intelligence Review. 2026.\n
\n\n
\n\n
\n\n
\n\n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{ArtificialReview25,\n\tauthor = {Yashar Mousavi and Parastoo Akbari and Rashin Mousavi and Arash Mousavi and Ibrahim Küçükdemiral and Afef Fekih and Ümit Çalı},\n\ttitle = {QRL-AFOFA: Q-Learning Enhanced Self-Adaptive Fractional Order Firefly Algorithm for Large-Scale and Dynamic Multiobjective Optimization Problems},\n\tjournal = {Artificial Intelligence Review},\n\tyear = {2026},\n\tdoi = {https://doi.org/10.1007/s10462-026-11511-y},\n\tkeywords = {J},\n}\n\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Data-Driven Pole Placement-Based $\\mathcal H_∞$ Controller Design via the Dilation Method in LMI Regions for Structural Vibration Control.\n \n \n \n\n\n \n Görmüş, B.; Yazıcı, H.; and Küçükdemiral, I. B.\n\n\n \n\n\n\n
Journal of Sound and Vibration, 362. 2026.\n
Under Review\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n\n\n\n
\n
@article{Bilal2026JSV, \n\ttitle = "Data-Driven Pole Placement-Based {$\\mathcal H_\\infty$} Controller Design via the Dilation Method in LMI Regions for Structural Vibration Control", \n\tauthor = "Bilal Görmüş and Hakan Yazıcı and Küçükdemiral, Ibrahim Beklan", \n\tyear = "2026", \n\tvolume = "362", \n\tjournal = "Journal of Sound and Vibration", \n\tpublisher = "Elsevier Ltd", \n\tnote = "Under Review",\n\tkeywords = "J",\n\t\n}\n\n\n\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Targeted cancer therapy by nonlinear data-enabled predictive control (DeePC): A case study on lewis lung carcinoma.\n \n \n \n\n\n \n Küçükdemiral, I. B.; Mousavi, Y.; and Busawon, K.\n\n\n \n\n\n\n
Biomedical Signal Processing and Control, 111: 108238. 2026.\n
\n\n
\n\n
\n\n
\n\n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{kucukdemiralBSPC26,\ntitle = {Targeted cancer therapy by nonlinear data-enabled predictive control (DeePC): A case study on lewis lung carcinoma},\njournal = {Biomedical Signal Processing and Control},\nvolume = {111},\npages = {108238},\nyear = {2026},\nissn = {1746-8094},\ndoi = {https://doi.org/10.1016/j.bspc.2025.108238},\nauthor = {Ibrahim Beklan Küçükdemiral and Yashar Mousavi and Krishna Busawon},\n}\n\n\n\n
\n\n\n\n
\n\n\n\n\n\n