Search

Optimally reconfiguring an active distribution network (ADN) during power outages has been regarded as a reasonable approach to facilitate system secure operation and reliability. Nevertheless, most existing studies for the reconfiguration virtually focus on taking actions from the generation- and network-side, in which the potential achievement from the demand-side is underestimated. Moreover, the phase-unbalance and voltage violation in ADNs should be restricted to avoid extreme conditions of distributed generators (DGs) that jeopardize system reliability. To bridge the gap, a new approach to reconfigure ADNs under multiple faults is proposed in this paper, incorporating a phase demand balancing (PDB) model to improve dispatch performance. The model regulates asymmetrical loads to mitigate the phase-unbalance issue from the demand-side, co-optimized with step voltage regulators (SVRs) and DG dispatching to enhance reliability and flexibility in reconfiguring ADNs. The derived optimization is a challenging mixed-integer non-convex programming (MINCP), which is reformulated as an efficiently solvable mixed-integer second-order cone programming (MISOCP) via exact equivalence and accurate approximation techniques. Case studies based on modified IEEE benchmark systems validate the effectiveness and advantages of the proposed method.

This letter proposes two novel effective transfer learning (TL) methods for power system stability assessment (SA) under distinct scenarios: cross-fault, where different types of faults are considered, and cross-scale, which accounts for varying system knowledge levels. Addressing the challenges faced in scenarios with few limited labeled SA data, our proposed data-driven SA models aim to transfer to the different but related scenarios by leveraging numerous instances from fully knowledge database and few labeled instances from the limited knowledge database. Moreover, a significant feature of our approach is the incorporation of the Extreme Learning Machine, a rapid neural network-based learning algorithm. Preliminary testing showcases an improvement of more than 24% in SA accuracy, especially for large-scale cross-scale transfer, demonstrating the efficacy of our TL techniques while maintaining computational efficiency.