Global Feature Enhancing and Fusion Framework for Strain Gauge Time Series Classification

TL;DR

The study tackles SGS recognition in IoT-enabled manufacturing where local time-series features from CNNs may fail to distinguish between subtly different patterns. It introduces the Global Feature Enhancing and Fusion (GFEF) framework, which combines feature-engineered global features (profile images and expert curvature metrics) with a hypergraph-based high-order fusion of multi-type features, aided by redundancy filtering and data reliability-aware attention. A Random Walk Hypergraph Construction and hypergraph attention-based information propagation learn and fuse cross-type global features, achieving superior accuracy on industrial SGS data and all 92 UCR datasets, with strong generalization and effective online deployment. The work demonstrates that high-order feature interactions and carefully engineered global features can significantly improve time-series classification in engineering contexts, enabling robust, scalable SGS monitoring and early fault warning.

Abstract

Strain Gauge Status (SGS) recognition is crucial in the field of intelligent manufacturing based on the Internet of Things, as accurate identification helps timely detection of failed mechanical components, avoiding accidents. The loading and unloading sequences generated by strain gauges can be identified through time series classification (TSC) algorithms. Recently, deep learning models, e.g., convolutional neural networks (CNNs) have shown remarkable success in the TSC task, as they can extract discriminative local features from the subsequences to identify the time series. However, we observe that only the local features may not be sufficient for expressing the time series, especially when the local sub-sequences between different time series are very similar, e.g., SGS data of aircraft wings in static strength experiments. Nevertheless, CNNs suffer from the limitation in extracting global features due to the nature of convolution operations. For extracting global features to more comprehensively represent the SGS time series, we propose two insights: (i) Constructing global features through feature engineering. (ii) Learning high-order relationships between local features to capture global features. To realize and utilize them, we propose a hypergraph-based global feature learning and fusion framework, which learns and fuses global features for semantic consistency to enhance the representation of SGS time series, thereby improving recognition accuracy. Our method designs are validated on industrial SGS and public UCR datasets, showing better generalization for unseen data in SGS recognition.

Figures (6)

• Figure 1: Time series strain gauge data (after normalization) of aircraft wing obtained from the web server. Experts point out the curvature of buckling curves is generally more pronounced than normal ones.
• Figure 2: The proposed Global Feature Enhancing and interaction Fusion (GFEF) framework with its components.
• Figure 3: Sorted average DRA scores on different datasets.
• Figure 4: Class activation map on the GunPoint dataset.
• Figure 5: Illustration of Shared Hypergraph Extractor of the proposed framework in Figure \ref{['fig:frame_work']}.