Celso Cavellucci, Fábio Luiz Usberti · Oct 08, 2026
Stochastic distribution network reconfiguration under demand uncertainty is investigated using a two-stage stochastic formulation, reducing expected losses by an average of 31.76% across 12 networks with varying topologies and demand scenarios. The reconfiguration also significantly reduces the number of voltage violations. Stochastic modeling only provides negligible additional benefit in this case due to the homothetic structure of the considered demand scenarios.
Why This Matters
This paper is highly relevant to power system engineers as it addresses the critical issue of distribution network reconfiguration under load uncertainty, which can significantly impact grid operations and resilience, particularly in systems with high levels of renewable energy integration. The results provide valuable insights for utility planners and grid operators seeking to optimize their networks and reduce expected losses while minimizing voltage violations.
Tien Dat Vu · Oct 08, 2026
A new framework for optimal control of unknown nonlinear systems uses predefined-time integral reinforcement learning to achieve stabilization in a specified time frame. The method involves approximating the unknown drift with a neural network and incorporating it into an IRL problem to learn a policy that converges within a prescribed deadline. The critic weights are updated using a novel update law, which ensures convergence within a specified residual set.
Why This Matters
This paper's focus on predefined-time integral reinforcement learning for optimal control of unknown nonlinear systems, specifically through the use of radial basis function neural networks and online identification laws, can be directly applicable to power system engineers in addressing issues such as grid stability, transient behavior, and adaptive control strategies, particularly in the context of integrating intermittent renewable energy sources.
Jie Fu, Anamika Dubey · Oct 08, 2026
A finite population of agents with decoupled Markov transition dynamics are subject to constraints ensuring that at least a fraction of agents reach a target region, while keeping the unsafe population below a certain threshold. To address this control problem, a method is proposed that incorporates second-order moment analysis and surrogate chance constraints to achieve tractable deterministic conditions. The approach is evaluated on two environments, including a gridworld and an EV-charging aggregation problem.
Why This Matters
This paper's focus on developing a method for policy synthesis under aggregate reach-avoid chance constraints has significant implications for power system engineers, particularly in the context of renewable integration and grid resilience. The proposed approach can help optimize charging patterns of electric vehicles to stabilize the grid and ensure reliability during peak demand periods.
Abanish Tiwari, Chandan Chaudhary, Mekh Raj Joshi et al. · Oct 07, 2026
A hybrid approach combining structural analysis based on centrality measures with a flow-based severity assessment is presented for rapid contingency screening in power systems. This method evaluates two indices: synergy index to quantify interactions among simultaneous line outages, and tolerance index to rank reinforcement candidates by their impact reduction. The approach was tested on the IEEE 39-bus system and identified 115 super-additive line pairs with significant interaction effects.
Why This Matters
This paper matters for power industry professionals as it presents a novel approach to rapidly screening contingencies and prioritizing transmission reinforcements, which can enhance grid resilience and reliability in the face of increasing complexity and variability. The results can inform ISO operations, utility planning, and FERC filings related to transmission system upgrades and capacity market design.
Energy Storage News