Abstract:Vibration data from physical systems, such as civil structures and machinery, often carries important information about the dynamic characteristics, but streaming acquisition of higher -frequency vibration often accrue large volumes of data, resulting in data transmission and storage challenges. Compressive sensing (CS) is a relatively newly-developed technique for efficient data representation, capable of reconstructing the target signal using only a few random measurements through sparse optimization. However, the real-world application of CS is hindered by the strong assumption of signal sparsity and a costly reconstruction process. In this work, we propose a novel deep learning method for vibration data reconstruction by using deep convolutional generative adversarial networks (DCGAN), which is composed of a generator G and a discriminator D. A modified 1D symmetric U-net architecture with shortcuts is presented for G to flexibly deal with different inputs, while a typical 1D classifier is used as D. A composite adversarial loss function is proposed considering errors in both time and frequency domains. The proposed DCGAN approach has several appealing properties. First, it directly learns the end-to-end mapping between the compressed and original signals without employing the sparsity assumption or random sampling, which fundamentally differs from existing sparsity -based CS methods. Second, the reconstruction process is highly computationally efficient as the network is fully feed-forward and no optimization is needed during data reconstruction. The proposed DCGAN approach is evaluated using the simulation data from a numerical 9-floor frame as well as experimental data collected from a large test steel grandstand. The results demonstrate the superiority of the proposed DCGAN in computational accuracy and efficiency compared to the tested sparsity-based algorithms. Furthermore, the influences of network configurations (network depth, down-sampling strategy, and shortcuts) are comprehensively explored.

Physics-informed Unsupervised Deep Learning Framework for Solving Full-Wave Inverse Scattering Problems

Deep-Learning Schemes for Full-Wave Nonlinear Inverse Scattering Problems.

Towards Intelligent Electromagnetic Inverse Scattering Using Deep Learning Techniques and Information Metasurfaces

Generalization Capabilities Of Deep Learning Schemes In Solving Inverse Scattering Problems

Enhanced Deep Learning Approach Based on the Deep Convolutional Encoder–Decoder Architecture for Electromagnetic Inverse Scattering Problems

Deep Learning-Based Inversion Methods for Solving Inverse Scattering Problems With Phaseless Data

A Fast Sparsity-Free Compressive Sensing Approach for Vibration Data Reconstruction Using Deep Convolutional GAN

Physics Embedded Deep Neural Network for Solving Full-Wave Inverse Scattering Problems

A Reconstruction Method of Electromagnetic Scattering Target Based on Diffusion Model

Electromagnetic Inverse Scattering With Perceptual Generative Adversarial Networks

Physics-Inspired Convolutional Neural Network for Solving Full-Wave Inverse Scattering Problems

Unrolled Optimization with Deep Learning-based Priors for Phaseless Inverse Scattering Problems

An Improved Deep Learning Scheme for Solving 2-D and 3-D Inverse Scattering Problems

Learning-Based Inversion Method for Solving Electromagnetic Inverse Scattering With Mixed Boundary Conditions

A Wavelet-Based Compressive Deep Learning Scheme for Inverse Scattering Problems.

Deep Injective Prior for Inverse Scattering

A More General Electromagnetic Inverse Scattering Method Based on Physics-Informed Neural Network

Multiple-Space Deep Learning Schemes for Inverse Scattering Problems

Electromagnetic Inverse Scattering with an Untrained SOM-Net

Nonlinear Electromagnetic Inverse Scattering Using DeepNIS

A Convolutional Sparsity Regularization for Solving Inverse Scattering Problems