PSI - Issue 52

ScienceDirect Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000 Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000 Available online at www.sciencedirect.com Procedia Structural Integrity 52 (2024) 203–213

www.elsevier.com/locate/procedia

www.elsevier.com/locate/procedia

2452-3216 © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of Professor Ferri Aliabadi 10.1016/j.prostr.2023.12.021 2452-3216 © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of Professor Ferri Aliabadi 2452-3216 © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of Professor Ferri Aliabadi © 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of Professor Ferri Aliabadi Abstract This paper presents a robust artificial neural network (ANN) loss function in solving samples with questionable data. The Latin hypercube sampling and the uniform experimental design are used as sample generation while ANN with various loss functions is the limit state function approaching method and Monte Carlo simulation is the failure probability computing technique. A marine lubricating oil cooler under a plus-minus triangular wave and internal pressure (1Mpa) is considered as representative of the dynamic implicit model. Three kinds of questionable data are considered: tiny vibrations( ± δ 1 ) caused by the surrounding environment, small deviations caused by equipment error( ± δ 2 ), and large deviations( ± Δ ) caused by accidental events. And two small probability events are considered: 1. Surrounding error + positive equipment error + a large shock ( ± Δ ) in partial samples; 2. Surrounding error + negative equipment error + a large shock ( ± Δ ) in partial samples. ANN with advanced loss function performs better than traditional loss function in dealing with samples including incorrect data. Keywords: Reliability analysis, ANN, MAPE, Robustness 1. Introduction Nowadays, more and more serving equipment needs to be evaluated and judged whether they can still fit the criterion in engineering. The increasing demand for reliability assessment in highly nonlinear structures and components among various subjects has fostered the research of low failure probability cases. Plenty of methods have been proposed to find the limit state function(LSF) and compute failure probability(P f ): Response surfaces, Polynomial Chaos Expansions, Support Vector Machines, Kriging and Artificial Neural Networks(ANN) (Teixeira R et al. Abstract This paper presents a robust artificial neural network (ANN) loss function in solving samples with questionable data. The Latin hypercube sampling and the uniform experimental design are used as sample generation while ANN with various loss functions is the limit state function approaching method and Monte Carlo simulation is the failure probability computing technique. A marine lubricating oil cooler under a plus-minus triangular wave and internal pressure (1Mpa) is considered as representative of the dynamic implicit model. Three kinds of questionable data are considered: tiny vibrations( ± δ 1 ) caused by the surrounding environment, small deviations caused by equipment error( ± δ 2 ), and large deviations( ± Δ ) caused by accidental events. And two small probability events are considered: 1. Surrounding error + positive equipment error + a large shock ( ± Δ ) in partial samples; 2. Surrounding error + negative equipment error + a large shock ( ± Δ ) in partial samples. ANN with advanced loss function performs better than traditional loss function in dealing with samples including incorrect data. Keywords: Reliability analysis, ANN, MAPE, Robustness 1. Introduction Nowadays, more and more serving equipment needs to be evaluated and judged whether they can still fit the criterion in engineering. The increasing demand for reliability assessment in highly nonlinear structures and components among various subjects has fostered the research of low failure probability cases. Plenty of methods have been proposed to find the limit state function(LSF) and compute failure probability(P f ): Response surfaces, Polynomial Chaos Expansions, Support Vector Machines, Kriging and Artificial Neural Networks(ANN) (Teixeira R et al. Fracture, Damage and Structural Health Monitoring Study on Robust Loss Function for Artificial Neural Networks Fracture, Damage and Structural Health Monitoring Study on Robust Loss Function for Artificial Neural Networks Models in Reliability Analysis Wu Zonghui a *, He Jian a , Sun Xiaodan a a College of Aerospace and Civil Engineering, Harbin Engineering University, Harbin 150001, China Models in Reliability Analysis Wu Zonghui a *, He Jian a , Sun Xiaodan a a College of Aerospace and Civil Engineering, Harbin Engineering University, Harbin 150001, China * Corresponding author. Tel.:+86 13386673015. E-mail address: wuzonghuib@hrbeu.edu.cn * Corresponding author. Tel.:+86 13386673015. E-mail address: wuzonghuib@hrbeu.edu.cn