An Integrated Method Based on Convolutional Neural Networks and Data Fusion for Assembled Structure State Recognition

This article focuses on the Assembled Structure (AS) state recognition method based vibration data. The difficulty of AS is mainly extraction effective classification features and pattern classification. paper presents an integrated Convolutional Neural Networks (CNNs) data fusion for recognition. takes wavelet transform time-frequency images denoised signal as input, uses CNNs to supervise learn data, extracts deep structure layer by layer, improves results through technology. tested assembly concrete shear wall using shake-table testing, show that it has a good overall identification accuracy (IA) 94.7%, indicating robust capable accurately recognizing very small changes in