A structure frequency response dynamic model correction method based on deep learning

Abstract

The invention discloses a structure frequency response dynamic model correction method based on deep learning. The method includes transforming the frequency response value of the experimental measured structure and the frequency response value of the dynamic model simulation into an image mode for storing and feature extraction, taking the corresponding parameters of the training sample images aslabels of the training set, and establishing the training sample sets of multi-frequency points, multi-observation points and multi-observation directions, and based on this, building a deep neural network and other processes. The method combines the advantages of deep learning in the field of image recognition, establishes the fast mapping relationship between the dynamic output and the parameters to be corrected, inputs the experimental measurement image into the trained neural network, outputs the model correction result, and effectively solves the problems of poor model representation ability of the manual feature extraction method and the like. In addition, considering the possibility of over-fitting caused by less training samples, the method of adding fast connection structure in the network forward transmission and adding noise expansion samples are adopted to reduce the parameter correction error.

Description

technical field [0001] The invention belongs to the field of structure dynamic frequency response dynamic model correction, relates to structural frequency response dynamic model correction and deep learning theory, and specifically relates to a structural frequency response dynamic model correction method based on deep learning. Background technique technical background: [0002] Structural dynamics models are often used in simulation experiments and computational analysis of complex structures. Due to various uncertainties and some simplifications in the modeling process, there is often a gap between the model and the actual structure. In order to improve the accuracy of the structural dynamics model, researchers have conducted a lot of research on the correction of the structural dynamics model. [1-4] . In the dynamic model correction method, there are two common correction methods based on modal data and frequency response function. Compared with the model correction...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is: to overcome the deficiency of multi-parameter, multi-measurement point, multi-frequency point, multi-directional frequency response data processing and artificial extraction feature method in the traditional method, the frequency response data of experimental measurement is transformed into multi- channel image

Images