This study evaluates four control strategies—rule-based, model predictive control (MPC), reinforcement learning without forecasts (RL‑NF), and reinforcement learning with forecasts (RL‑F)—for a renewable‑powered hydrogen supply chain. Using a unified, physically realistic simulation that includes electrolyzer constraints, storage dynamics, and grid limits, the authors find that MPC delivers the best economic performance by leveraging short‑term forecasts, while RL‑NF performs robustly without future information. RL‑F does not consistently outperform RL‑NF, indicating that forecast uncertainty and added state complexity can hinder forecast‑augmented learning.
By Mahammad Valiyev
The thesis presents a domain‑specific automated feature engineering pipeline, AutoEnergy, for energy consumption forecasting. AutoEnergy generates interpretable features from timestamps and lagged consumption, integrates with AutoML, and achieves 19.52%–84.72% error reductions across eighteen real‑world datasets while running 1.31–4.41 times faster than baseline methods. When combined with Decision‑Focused Learning for battery storage, it cuts operating costs by 22.9%–56.5% on a UK property dataset.
By Nasser Alkhulaifi
arXiv:2608. 02088v1 Announce Type: new Abstract: Reliable photovoltaic (PV) forecasts are needed for low-carbon energy systems, but newly deployed sites often have short, imperfect records.
By Fariba Dehghan, Sebastian Stein, Vahid Yazdanpanah, Stephanie Gauthier, Masood Nazari
The paper benchmarks a range of AI methods—conventional ML, ensemble learning, deep neural networks, recurrent architectures, Transformers, graph models, and hybrid ensemble deep learning—on three renewable energy datasets, including large‑scale wave energy converter (WEC) data and wind farm SCADA measurements. Tree ensembles, particularly Extra Trees, outperform traditional ML and neural predictors on structured WEC layout data, achieving a 63.7% MAE reduction over an MLP baseline. Spatial‑temporal graph networks (STGCN) and an RF‑BiLSTM hybrid further improve forecasting accuracy, with the hybrid model reaching an MAE of 150.5 kW, a 75% reduction over a standalone LSTM and 10% better than STGCN. The study concludes that no single architecture dominates; randomized ensembles excel for structured surrogate modeling, graph networks for explicit spatial interactions, and hybrid recurrent ensembles for combined nonlinear tabular and temporal dynamics.
arXiv:2511. 09789v3 Announce Type: replace Abstract: Short-term energy forecasting plays an important role in real-time operational decision-making, such as electricity market bidding and power system dispatch, where both numerical accuracy and correct directional signals are essential.
By Fulong Yao, Wanqing Zhao, Chao Zheng, Xiaofei Han
The paper introduces IMPACT, a physics‑informed multimodal deep‑learning framework that forecasts photovoltaic power 10–110 minutes ahead with probabilistic outputs. It uses a four‑branch architecture to encode plant‑level temporal history, inverter‑level spatial cloud dynamics, known‑future solar geometry, and current meteorological data, fusing them via gated cross‑attention. The model outperforms 14 baselines, achieving a forecast skill score of 0.2950 and maintaining strong performance under overcast and variable sky conditions.
By Anushka Katiyar, Cong Feng, Jing Wang, Tristan Liu, Simon Julien, Zachary Jacobs
Engineering use of AI forecasting models requires not only high nominal accuracy but also predictable behavior under uncertain inputs. In photovoltaic (PV) forecasting, this requirement is especially challenging because numerical weather prediction (NWP) errors are temporally correlated, state dependent, and physically coupled across variables.
The paper benchmarks a range of AI methods—conventional ML, ensemble learning, deep neural networks, recurrent architectures, Transformers, graph models, and hybrid ensemble deep learning—on renewable energy datasets. Tree ensembles, especially Extra Trees, outperform other models on structured wave energy converter (WEC) layout data, reducing MAE by about 63.7% relative to an MLP baseline. The RF BiLSTM hybrid delivers the best overall forecasting accuracy for wind farm SCADA data, cutting MAE by roughly 75% compared to a standalone LSTM and surpassing STGCN by about 10%.
By Majid Masoumi, Asghar Dashtiy, Mohammad Dehghan, Mina Rajabi
arXiv:2607. 12954v1 Announce Type: cross Abstract: Engineering use of AI forecasting models requires not only high nominal accuracy but also predictable behavior under uncertain inputs.
By Dandan Chen, Yan Zhao, Xuepeng Chen
arXiv:2606. 19199v1 Announce Type: cross Abstract: The recent growth of EV adoption poses challenges for power systems, including increased peak demand and potential grid instability.
By Giuseppe Gabriele, Fabio Pavirani, Seyed Soroush Karimi Madahi, Chris Develder
arXiv:2610.27155v1 Announce Type: new
Abstract: Accurate multi-horizon photovoltaic (PV) power forecasting is essential for grid stability, energy scheduling, and the efficient operation of PVbattery...
By Shafi Bin Sultan, Sabik Bin Sultan, Mohammad Tajrian Arafat Alvi, Rayhanul Islam Sony, Abdulah Al Numan
The paper presents a machine‑learning based climate classification tailored for photovoltaic (PV) modules, incorporating both energy yield and module lifetime with climate‑dependent degradation. Using an interpolated dataset of twelve input features, the authors identify annual global horizontal irradiation and ambient temperature as the most influential predictors, achieving RMSEs of 0.007 MWh for yield and 1.5 years for lifetime. The resulting hierarchical clustering yields six primary climate clusters (Tropical, Desert, Continental, Temperate, Boreal, Polar) and 15 subclusters, with the low‑temperature continental climate delivering the highest discounted lifetime energy yield.
By Youri Blom, Sofia Dutto, Alexandru Costache, Rowan Richie, Ruben Pelsser, Wesley Berger, Jing Sun, Rudi Santbergen, Olindo Isabella, Malte Ruben Vogt