Deep learning detects electrical faults in aircraft power systems
Deep Learning-Based Detection of Electrical Faults and Power Quality Disturbances in Aerospace Power Systems
Machine Learning
Summary
Electrical systems in airplanes operate at high frequencies that differ from typical power grids, making it hard to spot problems fast. The authors created a computer program using deep learning that can identify various electrical faults and quality issues in an aircraft’s 400 Hz electrical supply. They trained and tested the program on simulated data based on a real airplane’s design and found it works accurately and runs quickly on specialized hardware. This approach could help monitor airplane electrical health more reliably in real time. The authors also shared data to help others build on their work.
What this means in practice
- •For aircraft electrical engineers: Implement real-time fault detection in aircraft power systems using a compact deep learning model verified on hardware.
- •For embedded system developers: Deploy efficient deep learning models for monitoring high-frequency electrical signals on edge hardware with minimal latency.
Authors
Ian C. Guzmán, Radu Babiceanu, Berker Peköz
Abstract
More Electric Aircraft require fast and reliable monitoring of high-frequency electrical networks, yet most power quality disturbance and fault diagnosis methods are developed for conventional 50 or 60 Hz grids. This work presents a hardware-aware deep learning framework for multiclass detection of electrical faults and power quality disturbances in a 400 Hz aerospace power system. A high-fidelity simulation model inspired by the Boeing 787 electrical architecture generates voltage and current waveforms for 21 normal, disturbance, switching, open-circuit, and short-circuit conditions. Two datasets, each containing 73,500 samples, are formed from one-dimensional time-series signals and short-time Fourier transform time-frequency representations. Signal-processing augmentation, domain randomization, and class-specific generative adversarial networks increase waveform diversity, and the time-series dataset is released through IEEE DataPort. We compare 1D and 2D convolutional neural networks, long short-term memory networks, CNN-LSTM hybrids, ResNet, MobileNet, and VGG models under common training conditions. A compact ResNet provides the best accuracy-complexity tradeoff, achieving 96.94 percent software test accuracy with 175,685 parameters. After 8-bit quantization and deployment on a Xilinx Zynq UltraScale Plus MPSoC ZCU102, the model achieves 95.87 percent accuracy and a measured mean neural-network accelerator latency of 6.90 ms per input record. The results establish simulation-based, accelerator-level feasibility for embedded edge AI in aircraft electrical health monitoring and motivate future end-to-end data acquisition and experimental validation.