Explainable Artificial Intelligence based Soft Evaluation Indicator for Arc Fault Diagnosis

2025-07-21

AI Summary

Existing AI-based arc-fault diagnosis models achieve high accuracy but suffer from poor interpretability, undermining output credibility. To address this, we first formulate formal interpretability criteria for arc-fault detection and propose a novel soft evaluation metric grounded in eXplainable AI (XAI). We further design a lightweight balanced neural network that jointly optimizes classification accuracy, model interpretability, and feature extraction capability. Our method is rigorously validated on a real-world arc-fault dataset featuring multiple noise levels and sampling rates, demonstrating significant improvements in model transparency and decision trustworthiness across two heterogeneous experimental scenarios—while maintaining state-of-the-art classification accuracy. The core contributions are threefold: (1) the first quantitative, domain-specific interpretability standard for arc-fault diagnosis; (2) a new XAI-aware evaluation framework that harmonizes accuracy and interpretability; and (3) an efficient neural architecture enabling synergistic optimization of both objectives.

Technology Category

• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Computer Vision: Interpretability, Explainability, and Transparency

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Algorithmic accountability and transparency on the web
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Novel AI-based arc fault diagnosis models have demonstrated outstanding performance in terms of classification accuracy. However, an inherent problem is whether these models can actually be trusted to find arc faults. In this light, this work proposes a soft evaluation indicator that explains the outputs of arc fault diagnosis models, by defining the the correct explanation of arc faults and leveraging Explainable Artificial Intelligence and real arc fault experiments. Meanwhile, a lightweight balanced neural network is proposed to guarantee competitive accuracy and soft feature extraction score. In our experiments, several traditional machine learning methods and deep learning methods across two arc fault datasets with different sample times and noise levels are utilized to test the effectiveness of the soft evaluation indicator. Through this approach, the arc fault diagnosis models are easy to understand and trust, allowing practitioners to make informed and trustworthy decisions.

Problem

• Develops explainable AI indicator for arc fault diagnosis trustworthiness
• Proposes lightweight neural network for accurate soft feature extraction
• Tests evaluation indicator on diverse datasets and noise levels

Innovation

• Explainable AI for arc fault diagnosis
• Lightweight balanced neural network
• Soft evaluation indicator for trust