Energy Digest

A new algorithm is proposed for forming coalitions in distributed energy systems with limited information sharing to address privacy concerns and reduce computational overhead. The method uses an upper bound on the value of candidate coalitions to eliminate optimization problems, significantly reducing complexity and limiting information exchange. The algorithm demonstrates improved economic performance over decentralized control with lower computational cost than full-information approaches.

A proposed start-up scheme for real-time emergency control in power systems uses three technologies: an instability index, a Critical Machines identification algorithm, and a two-layer Single Machine Infinite Bus equivalence framework. The scheme shows good accuracy and high reliability, can identify transient instability accurately and fast, and is robust to measurement errors. It significantly reduces communication burden and improves computation efficiency compared to conventional methods.

The paper proposes an optimal additional active power control (AAPC) of wind turbines to improve grid frequency stability by maximizing the frequency nadir post a major power deficit. The proposed method decouples WT response and frequency dynamics, solving the optimal frequency trajectory and constructing the optimal AAPC based on this trajectory. Simulation results show that the proposed method can significantly improve the system frequency nadir.

Researchers developed a framework to analyze the Age of Incorrect Information in real-time monitoring systems using discrete-time Markov sources, introducing a novel multiple-threshold policy class for optimizing transmission rates and reducing computational complexity. The proposed policies outperform periodic scheduling in numerical experiments, achieving performance comparable to the globally optimal policy. A simplified single-threshold policy can be derived at the cost of optimality.

A communication and control strategy is proposed for stabilizing continuous-time linear systems under finite data-rate constraints, using sampled and quantized state measurements with dynamically adjusted quantization ranges to handle escapes from the quantization range. The strategy guarantees input-to-state stability (ISS) and employs an additional quantization symbol to ensure robustness near equilibrium. It improves upon existing results that yield only practical ISS or lack explicit data-rate conditions.

Researchers propose an optimal, carbon-aware optimization strategy that stores solar surplus energy by adjusting indoor temperature setpoint within comfort bounds, reducing emissions and grid electricity consumption in buildings. This approach avoids dedicated batteries, addressing environmental and cost concerns, while maintaining comfort levels. The strategy is evaluated through simulations of three TRNSYS models, showing consistent reductions in energy consumption compared to a baseline scenario.

A computationally tractable framework for predictive control of multi-chiller plants has been developed, extending Differentiable Predictive Control to accommodate mixed-integer control policies. The proposed approach achieves significant energy savings compared to a fast heuristic Rule-Based Controller while maintaining faster computation times than Model Predictive Control. It offers a scalable alternative to conventional combinatorial mixed-integer control formulations.