工艺与制造英文2021被引 6
自动纤维铺放工艺制造缺陷光学检测中深度学习分类器决策的可解释性
Explainability of deep learning classifier decisions for optical detection of manufacturing defects in the automated fiber placement process
Sebastian Meister, Mahdieu Wermes, Jan Stüve, Roger M. Groves · Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR)
摘要整理
自动纤维铺放是航空航天领域常用的复合材料制造工艺，传统上需要人工目视检测。神经网络缺陷分类具有自动化目视检测的潜力，但机器决策过程难以验证。本研究提出了一种卷积神经网络（CNN）制造缺陷分类的可视化方法，并量化了其鲁棒性。研究表明，平滑积分梯度法（Smoothed Integrated Gradients）和DeepSHAP特别适合CNN分类的可视化。平滑积分梯度法在评估降质输入图像时还表现出更好的鲁棒性优势。该方法通过可解释性技术增强了自动纤维铺放过程中缺陷检测的可信度，为实现工业级自动化检测系统奠定了基础。
相关论文
A review of recent developments in natural fibre composites and their mechanical performance
Composites Part A Applied Science and Manufacturing · 2015
Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: Basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee
Clinical Neurophysiology · 2015
Lignocellulosic biomass: a sustainable platform for the production of bio-based chemicals and polymers
Polymer Chemistry · 2015
Design for Additive Manufacturing: Trends, opportunities, considerations, and constraints
CIRP Annals · 2016
Nanoimprint Lithography: Methods and Material Requirements
Advanced Materials · 2007
Nanocomposites: synthesis, structure, properties and new application opportunities
Materials Research · 2009
← 返回论文库整理：复材站编辑部