Latency-Aware Deep Learning Benchmark for Real-Time Cyber-Physical Attack and Fault Classification in Inverter-Dominated Power Grids

TL;DR

This paper presents a latency-aware benchmark for deep learning models in power system anomaly detection, highlighting the gap between current inference latency and real-time protection requirements.

Contribution

It introduces a reproducible benchmarking framework and evaluates multiple neural network architectures on high-fidelity power system data for real-time fault and attack classification.

Findings

All models classified events within one cycle (<15 ms)

End-to-end latency ranged from 50 to 90 ms, exceeding real-time requirements

Highlights need for optimization and hardware acceleration

Abstract

This work introduces a latency-aware benchmarking framework for evaluating deep learning models in power system anomaly detection using high-fidelity, time-domain signals generated from an industry-grade electromagnetic transient simulator. Eight neural network architectures, ranging from MLPs to Transformers, were systematically evaluated on streaming datasets representing both physical faults and cyber-attacks in inverter-dominated networks. All models successfully classified two representative multi-event sequences in real time with sub-cycle response times below 15 ms. However, although classification decisions occurred within one cycle, the end-to-end inference latency consistently exceeded three cycles, ranging from 50 to 90 ms. These results highlight a critical gap between algorithmic capability and protection-grade deployment, pointing to the need for further optimization and hardware acceleration. The findings establish a reproducible benchmark for sub-cycle anomaly detection and provide guidance for transitioning machine learning methods from research prototypes to real-world protection applications.