Energy Digest

Closed-loop stability of MPC and infinite-horizon optimal control can be guaranteed under certain conditions, including continuity of the model and cost-controllability. Exponential stability is possible with exponential decay rates that depend on the size of plant-model mismatch bounds proportional to states and controls. A suboptimality bound exists for the closed-loop cost, recovering a portion of the optimal cost of the surrogate model.

The article proposes a novel data-driven power flow framework for radial distribution networks, combining behavioral and DistFlow models to solve power flow problems using limited real-time measurements, enabling accurate voltage magnitude predictions with minimal sensor placement. The proposed algorithm is designed for practical distribution grids with sparse measurement availability, reducing the number of required sensors. It demonstrates accuracy in several test cases with a maximum error less than 0.001 p.u.

A novel generator torque controller regulates rotor speed for floating offshore wind farms, specifically in regions 2 and 3 with different control strategies. Two modeling paradigms, NL-MPC and LPVTD-MPC, achieve high power tracking scores above 89.9% in three testing scenarios, but LPVTD-MPC solves the problem faster but with larger fluctuations. The framework is a pioneering approach to multi-region wind turbine dynamics for floating offshore wind farms.

A game-theoretic decentralized real-time control of electric vehicle charging stations is proposed using the Stackelberg Game-based Alternating Direction of Multipliers (SG-ADMM) algorithm. The approach integrates SG-ADMM into an Energy Management System to optimize EV power dispatch in a distributed manner, incentivizing EVs to align with the charging station's optimum goals. The method outperforms other approaches in terms of cost-effectiveness, fairness, and computational efficiency.

Decentralized control systems are being explored to manage large-scale Electric Vehicle Charging Stations due to limitations in centralized approaches. A game-theoretic approach is used to design an incentive mechanism that aligns individual EV objectives with the controller's optimum, employing a Stackelberg Game-based Alternating Direction Method of Multipliers (SG-ADMM) algorithm. This approach aims to optimize EV charging management in a distributed manner.

Estimating generation costs in electricity markets can be achieved through Bayesian inference using a latent variable model and balanced neural posterior estimation (BNPE). This approach recovers marginal costs with narrow credible intervals, while start-up costs remain largely unidentifiable from schedules alone. The method outperforms an inverse-optimization algorithm that exhibits larger parameter errors without uncertainty quantification.

Harnessing demand-side flexibility in building and mobility sectors can help integrate renewable energy into power systems, reducing global CO2 emissions. This sector coupling can be achieved through advances in energy management, business models, control technologies, and power grids, requiring deep integration across stakeholders and systems. Fully realizing demand-side flexibility necessitates transdisciplinary collaboration and commercially viable solutions to maximize its multi-stakeholder benefits.

Aggregated single-unit doubly-fed induction generator (DFIG) models show different small-signal stability boundaries compared to their granular counterparts. This disparity can lead to misleading stability assessments due to the aggregation effect on D-decomposition-related characteristics. Stability boundaries differ in shape, critical modes, and evolution patterns between aggregated and granular models under various control settings and operating conditions.