A hybrid ANN-based imperial competitive algorithm methodology for structural damage identification of slab-on-girder bridge using data mining

Abstract

Implementation of data mining (DM) techniques in different areas of civil engineering has recently given very good results. However, application of DM in structural health monitoring (SHM) is not used as much as expected, thus, many challenges are still ahead. Therefore, it seems a vital need is required to develop the applicability of DM in SHM. To this end, the current study attempts to present a DM-based damage detection methodology using modal parameter data, which trained by means of a hybrid artificial neural network-based imperial competitive algorithm (ANN-ICA). Likewise, the hybrid ANN is optimized by a new optimization-based evolutionary algorithm, called ICA, to predict the severity and location of multiple damage cases obtained from experimental modal analysis of intact and damaged slab-on-girder bridge structures. Furthermore, the applicability of DM approach was developed to detect the hidden patterns in vibration data using Cross Industry Standard Process for DM (CRISP-DM) tool. The performance of the model was carried out using comparison of a pre-developed ANN and ANN-ICA model.