More Like this

1. A Meta-Strategy Framework With Deep Reinforcement Learning for Damping Interarea Oscillation via High-Voltage DC Transmission

   https://doi.org/10.1109/TII.2025.3594123  

   Gao, Wei; Fan, Rui; Huang, Qiuhua; Li, Yushuai; Gao, David Wenzhong (November 2025, IEEE Transactions on Industrial Informatics)

   Inter‑area oscillations pose significant threats to power‑system security and stability, requiring rapid damping to prevent potential system failure. Our previous work introduced a deep reinforcement learning (DRL) strategy built on the power oscillation damping (POD) control architecture using high‑voltage direct current (HVDC) transmissions, which effectively generated POD signals from wide‑area information to modulate transferred power and damp inter‑area oscillations. However, its performance degrades under significant structural or operating‑condition changes, revealing a general adaptability limitation in DRL‑based control schemes. To address this issue, we propose a novel meta evolution strategy (MES) framework that integrates meta‑strategy optimization (MSO) with evolutionary‑strategy (ES)‑based DRL. The ES algorithm uses fewer, easier‑to‑tune hyperparameters than typical DRL methods and rapidly identifies optimal policies by exploring the policy parameter space through parallel computation, while MSO introduces a latent variable that projects power‑system parameters under varying operating conditions to enable fast, efficient adaptation to unforeseen scenarios. A modified IEEE 39‑bus system with HVDC transmission is studied to validate the performance and effectiveness of the proposed meta‑strategy in damping inter‑area oscillations. 

   more » « less
2. Deep reinforcement learning based control for enhanced frequency response with multi-energy storage systems

   https://doi.org/10.1016/j.epsr.2025.112432  

   Hasan, Abu Shouaib; Fan, Rui; Gao, Wei; Wu, Di (January 2026, Electric Power Systems Research)

   Not AvailaThis paper proposes an advanced strategy for managing multiple battery energy storage systems (BESS) to enhance frequency support during contingencies. A novel deep reinforcement learning (DRL) framework based on a guided surrogate-gradient-based evolutionary strategy (GSES) was developed to dynamically regulate BESS outputs for rapid power injection or absorption. This approach effectively mitigates the rate of change of frequency (ROCOF) and stabilizes the system frequency under varying operating conditions. Parallel computing techniques are employed to accelerate training and ensure robust performance. In addition, a genetic algorithm is implemented to determine the placement of BESS within the grid network, strategically minimizing ROCOF during disturbances by accounting for active power injections and inertia contributions from multiple BESS units. The proposed GSES-DRL methodology is rigorously validated using extensive simulations on a modified IEEE 39-bus system, demonstrating effective performance in frequency response compared with existing strategies. In severe scenarios, the proposed method successfully prevents unnecessary load-shedding relay operations, thereby enhancing system stability. 

   more » « less
3. A Deep Long Short-Term Memory Network for Bearing Fault Diagnosis Under Time-Varying Conditions

   https://doi.org/10.1115/DETC2022-88808  

   Zhou, Kai (August 2022, ASME 2022 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference)

   Abstract The fault diagnosis of bearing in machinery system plays a vital role in ensuring the normal operating performance of system. Machine learning-based fault diagnosis using vibration measurement recently has become a prevailing approach, which aims at identifying the fault through exploring the correlation between the measurement and respective fault. Nevertheless, such correlation will become very complex for the practical scenario where the system is operated under time-varying conditions. To fulfill the reliable bearing fault diagnosis under time-varying condition, this study presents a tailored deep learning model, so called deep long short-term memory (LSTM) network. By fully exploiting the strength of this model in characterizing the temporal dependence of time-series vibration measurement, the negative consequence of time-varying conditions can be minimized, thereby improving the diagnosis performance. The published bearing dataset with various time-varying operating speeds is utilized in case illustrations to validate the effectiveness of proposed methodology. 

   more » « less
4. A Study of Linear Programming and Reinforcement Learning for One-Shot Game in Smart Grid Security

   https://doi.org/10.1109/IJCNN.2018.8489202  

   Paul, Shuva; Ni, Zhen (July 2018, 2018 International Joint Conference on Neural Networks (IJCNN))

   Smart grid attacks can be applied on a single component or multiple components. The corresponding defense strategies are totally different. In this paper, we investigate the solutions (e.g., linear programming and reinforcement learning) for one-shot game between the attacker and defender in smart power systems. We designed one-shot game with multi-line- switching attack and solved it using linear programming. We also designed the game with single-line-switching attack and solved it using reinforcement learning. The pay-off and utility/reward of the game is calculated based on the generation loss due to initiated attack by the attacker. Defender's defense action is considered while evaluating the pay-off from attacker's and defender's action. The linear programming based solution gives the probability of choosing best attack actions against different defense actions. The reinforcement learning based solution gives the optimal action to take under selected defense action. The proposed game is demonstrated on 6 bus system and IEEE 30 bus system and optimal solutions are analyzed. 

   more » « less
5. Feature Corrective Transfer Learning: End-to-End Solutions to Object Detection in Non-Ideal Visual Conditions

   https://doi.org/10.1109/cvprw63382.2024.00007  

   Wei, Chuheng; Wu, Guoyuan; Barth, Matthew J (June 2024, IEEE)

   A significant challenge in the field of object detection lies in the system's performance under non-ideal imaging conditions, such as rain, fog, low illumination, or raw Bayer images that lack ISP processing. Our study introduces "Feature Corrective Transfer Learning", a novel approach that leverages transfer learning and a bespoke loss function to facilitate the end-to-end detection of objects in these challenging scenarios without the need to convert non-ideal images into their RGB counterparts. In our methodology, we initially train a comprehensive model on a pristine RGB image dataset. Subsequently, non-ideal images are processed by comparing their feature maps against those from the initial ideal RGB model. This comparison employs the Extended Area Novel Structural Discrepancy Loss (EANSDL), a novel loss function designed to quantify similarities and integrate them into the detection loss. This approach refines the model's ability to perform object detection across varying conditions through direct feature map correction, encapsulating the essence of Feature Corrective Transfer Learning. Experimental validation on variants of the KITTI dataset demonstrates a significant improvement in mean Average Precision (mAP), resulting in a 3.8-8.1% relative enhancement in detection under non-ideal conditions compared to the baseline model, and a less marginal performance difference within 1.3% of the mAP@[0.5:0.95] achieved under ideal conditions by the standard Faster RCNN algorithm. 

   more » « less