Structural damage identification based on autoencoder neural networks and deep learning. (1st October 2018)

• Record Type: Journal Article
• Authors: Pathirage, Chathurdara Sri Nadith; Li, Jun; Li, Ling; Hao, Hong; Liu, Wanquan; Ni, Pinghe
• Is Part Of: Engineering structures. Volume 172(2018)
• Page Start: 13
• Page End: 28
• Publication Date: 2018-10-01
• Subjects: Autoencoders -- Deep learning -- Deep neural networks -- Structural damage identification -- Pre-training; Structural engineering -- Periodicals; Structural analysis (Engineering) -- Periodicals; Construction, Technique de la -- Périodiques; Génie parasismique -- Périodiques; Pression du vent -- Périodiques; Earthquake engineering; Structural engineering; Wind-pressure; Periodicals; 624.105
• DOI: 10.1016/j.engstruct.2018.05.109
• Languages: English
• ISSNs: 0141-0296

Abstract

Highlights: An autoencoder based framework for structural damage identification is proposed. It supports deep neural networks for solution of pattern recognition problems. Dimensionality reduction and relationship learning are included in the framework. Accurate identification results are obtained considering uncertainties and noises. Numerical and experimental investigations are conducted to validate the approach. Abstract: Artificial neural networks are computational approaches based on machine learning to learn and make predictions based on data, and have been applied successfully in diverse applications including structural health monitoring in civil engineering. It is difficult to optimize the weights in the neural networks that have multiple hidden layers due to the vanishing gradient issue. This paper proposes an autoencoder based framework for structural damage identification, which can support deep neural networks and be utilized to obtain optimal solutions for pattern recognition problems of highly non-linear nature, such as learning a mapping between the vibration characteristics and structural damage. Two main components are defined in the proposed framework, namely, dimensionality reduction and relationship learning. The first component is to reduce the dimensionality of the original input vector while preserving the required necessary information, and the second component is to perform the relationship learning between the features with the reduced dimensionality and the stiffness reduction parameters of the structure. Vibration characteristics, such as natural frequencies and mode shapes, are used as the input and the structural damage are considered as the output vector. A pre-training scheme is performed to train the hidden layers in the autoencoders layer by layer, and fine tuning is conducted to optimize the whole network. Numerical and experimental investigations on steel frame structures are conducted to demonstrate the accuracy and efficiency of the proposed framework, comparing with the traditional ANN methods.