Transferable Neuromorphic Protection for Distribution Networks

Protection of converter-dominated distribution networks is challenging due to limited fault currents, operating-mode transitions, bidirectional power flow, and heterogeneous network configurations. This paper proposes a scale-invariant neuromorphic protection framework to enable transferable protection decisions across such networks. The framework integrates asset-based normalization, multi-domain disturbance representation, event-driven spike encoding, leaky integrate-and-fire (LIF) dynamics, adaptive threshold regulation, and directional supervision within a unified mathematical formulation. Local voltage, current, active power, and reactive power measurements are transformed into dimensionless disturbance representations, while latency encoding and membrane integration capture disturbance intensity and temporal evolution. The framework is evaluated across four heterogeneous network configurations: a two-DER inverter-interfaced microgrid, an islanded ring microgrid, the IEEE 13-node feeder, and the CIGRE European medium-voltage benchmark. Across 480 cases, it achieves a median operating time of 7.6 ms, with 94.2 % of cases operating within one power cycle. The results demonstrate selective protection across different electrical ratings and network topologies using a common parameterization without network-specific retuning, highlighting the potential of neuromorphic processing for physically interpretable, communication-free protection of converter-dominated distribution networks.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-12
DOI
https://doi.org/10.48550/arxiv.2611.16702
Primary Topic
Power Systems Fault Detection
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Transferable Neuromorphic Protection for Distribution Networks

Subham Sahoo, Saurabh Prabhakar, Bijaya Ketan Panigrahi, Yubo Song et al.
arXiv (Cornell University)
Power Systems Fault Detection
preprint

Transferable Neuromorphic Protection for Distribution Networks

Subham Sahoo, Saurabh Prabhakar, Bijaya Ketan Panigrahi, Yubo Song, Pascal Schirmer
preprint en

Abstract

Protection of converter-dominated distribution networks is challenging due to limited fault currents, operating-mode transitions, bidirectional power flow, and heterogeneous network configurations. This paper proposes a scale-invariant neuromorphic protection framework to enable transferable protection decisions across such networks. The framework integrates asset-based normalization, multi-domain disturbance representation, event-driven spike encoding, leaky integrate-and-fire (LIF) dynamics, adaptive threshold regulation, and directional supervision within a unified mathematical formulation. Local voltage, current, active power, and reactive power measurements are transformed into dimensionless disturbance representations, while latency encoding and membrane integration capture disturbance intensity and temporal evolution. The framework is evaluated across four heterogeneous network configurations: a two-DER inverter-interfaced microgrid, an islanded ring microgrid, the IEEE 13-node feeder, and the CIGRE European medium-voltage benchmark. Across 480 cases, it achieves a median operating time of 7.6 ms, with 94.2 % of cases operating within one power cycle. The results demonstrate selective protection across different electrical ratings and network topologies using a common parameterization without network-specific retuning, highlighting the potential of neuromorphic processing for physically interpretable, communication-free protection of converter-dominated distribution networks.

arXiv (Cornell University)
Power Systems Fault Detection
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon