Damage detection in steel plates using feed-forward neural network coupled with hybrid particle swarm optimization and gravitational search algorithm

Abstract

Over recent decades, the artificial neural networks (ANNs) have been applied as an effective approach for detecting damage in construction materials. However, to achieve a superior result of defect identification, they have to overcome some shortcomings, for instance slow convergence or stagnancy in local minima. Therefore, optimization algorithms with a global search ability are used to enhance ANNs, i.e. to increase the rate of convergence and to reach a global minimum. This paper introduces a two-stage approach for failure identification in a steel beam. In the first step, the presence of defects and their positions are identified by modal indices. In the second step, a feedforward neural network, improved by a hybrid particle swarm optimization and gravitational search algorithm, namely FNN-PSOGSA, is used to quantify the severity of damage. Finite element (FE) models of the beam for two damage scenarios are used to certify the accuracy and reliability of the proposed method. For comparison, a traditional ANN is also used to estimate the severity of the damage. The obtained results prove that the proposed approach can be used effectively for damage detection and quantification.

An improved Artificial Rabbit Optimization for structural damage identification

“…By incorporating these parameters into the objective function of the metaheuristic algorithm, the algorithm searches for the optimal combination of damage locations and severities that best match the observed vibration data such as Particle Swarm Optimization (PSO) (Gökdağ and Yildiz, 2012;Kang et al, 2012;Nguyen-Ngoc et al, 2021); Teaching Learning Based Optimization (TLBO) (Ahmadi-Nedushan and Fathnejat, 2022;Shahrouzi and Sabzi, 2018); Slime Mould Algorithm (SMA) (Ngoc-Nguyen et al, 2023), Artificial Hummingbird Algorithm (AHA) ; Marine Predator Algorithm (MPA);…Besides, many hybrid metaheuristic algorithms combined with neural networks (Hoàng Việt et al, 2023;Su Fen et al, 2023) have also been applied to fault diagnosis, Feedforward Neural Networks and Marine Predator Algorithm (MPAFNN) (Ho et al, 2021b); Grey Wolf Optimizer and Artificial Neural Networks (GWOANN) (Ho Viet et al, 2022); Particle Swarm Optimization-Gravitational Search Algorithm (PSOGSA); Feedforward Neural Network-Particle Swarm Optimization and Gravitational Search Algorithm (FNN-PSOGSA) (Ho et al, 2021a).…”

An improved version of mode shape based indicator for structural damage identification

“…Another approach is to use a damage index derived from frequencies and/or mode shapes, which can directly indicate damage location without an iterative process [9][10][11][12][13]. These studies share the same assumption that information about the pristine and damaged structure is already achieved.…”

An efficient stochastic-based coupled model for damage identification in plate structures

“…Although all these mentioned studies, in general, showed high potential in damage localization, they did not indicate a quantitative relationship between the damage indices and the reduction in stiffness. To solve this problem, many authors combined damage indicators with an optimization process [23][24][25][26] or used a hybrid model between optimization and feedforward neural network (FNN) coupled with damage indicator [27,28]. The first approach identifies the damage by solving inverse problems or updating some model parameters until meeting a superior agreement between the FE model and the measured modal parameters.…”