Nondestructive Evaluation of Ducts in Prestressed Concrete Bridges Using Heterogeneous Neural Networks and Impact-echo

Abstract

Prestressed concrete (PSC) girder bridges are widely used owing to their economic efficiency, durability, and effective maintenance. However, since voids in ducts may cause sudden structural collapse, it is very important to detect them early. To solve this problem, voids are detected by analyzing the impact-echo (IE) signal measured by IE equipment containing a sensor, but it is difficult to accurately detect voids in a short time even by experts. In this study, we aim to detect voids in ducts on the basis of various types of neural networks and IE signals. For more effective learning, the raw IE signal is filtered and then used in its specific range, and it is also converted into a frequency spectrum by the Fourier transform. The filtered IE signal is trained with long short-term memory (LSTM) to reflect the characteristics of its time series. The frequency spectrum is trained with a feed-forward neural network because it is not a time series. After that, a multiplication operation is performed on the outputs of each network, and a model capable of detecting the internal voids of ducts is created by training these integrated features. In the experimental results, our proposed model showed an accuracy of 97.474%.Yu-Seop Kim received his Ph.D. degree in computer engineering from Seoul National University, Republic of Korea. He is currently a professor in the School of Software, Hallym University. His research interests are in the areas of bioinformatics, computational intelligence, and natural language processing.

Void Detection inside Duct of Prestressed Concrete Bridges Based on Deep Support Vector Data Description

“…It performs training and evaluation using just Specimen−2 data divided in an 8:1:1 (train:validation:test) ratio. ( 2) With IE data collected in different environments, compare the performance of the supervised learning model [26][27][28] with our proposed models. It performs an evaluation using the IE data in Specimen−2 and trains using the IE data in Specimen−1.…”

“…In the first experiment (using the IEs collected in the same specimen), we trained and evaluated the performance of the supervised learning model developed in the previous studies [26][27][28]. The results are shown in Table 3.…”

“…Detecting voids that arise inside the ducts of PSC bridges is a particularly significant issue; however, it has only been investigated using a signal processing-based method, and machine learning is rarely applied. Therefore, by utilizing IE data and LSTM, supervised learning was attempted to detect defects inside ducts in previous studies [26][27][28]. Utilizing IE data and structural information (concrete thickness, depth of duct, distance between the measured point and impact point), [26] used LSTM to classify the duct's voids.…”

“…However, we determined that the structural information affects the scalability of the deep learning-based approach. Therefore, [27,28] developed an LSTM-based model that can use a frequency spectrum with IE data.…”