Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Design-focused Interpretable Machine Learning Models for Compressive Capacity Prediction of Gusset Plate Connections
https://orcid.org/0000-0001-8012-8925
https://orcid.org/0000-0002-2561-993X
Highlights An interpretable machine learning framework is developed to design gusset plate connection. A comprehensive dataset consisting of experimental tests and numerical simulation is assembled. Boosting algorithms (XGBoost, CATBoost) demonstrated very high accuracy in predicting the gusset plate capacity and outperformed existing empirical models. Feature importance as well as the global and local explanations of the developed models are explained employing SHAP. A reliability analysis is performed to identify the level of safety of predictions obtained by the CatBoost and XGBoost models.
Abstract Gusset plates can experience large compressive loads when used to connect bracing systems to beam-column joints in steel frames and truss bridges. The strength of a gusset plate is therefore vital to allow the bracing members to reach their ultimate capacity without a brittle failure of the gusset plate connection. Due to the large complexity of the stress distribution mechanism, simplified approaches commonly adopted for design can lead to inaccurate and highly conservative results. The present study aims at developing a simple yet efficient gusset plate compressive strength prediction model that does neither rely on crude assumptions nor require complicated calculations. A meticulously compiled database consisting of 68 experimental and 184 numerical datasets was deployed to aid in developing a high-performance machine learning (ML) model selected amongst 10 machine learning techniques studied. The superiority of the CatBoost and XGBoost models was confirmed with an accuracy of 95% and 96%, respectively significantly outperforming current design practices and widely used empirical formulas. SHapley Additive exPlanations (SHAP) and Taylor diagram were used to explain the feature importance and model performance. The developed interpretable ML models can instill trust in a model that explains its decisions, as opposed to other black box approaches. It was found that CatBoost was the most efficient in predicting the compressive strength of gusset plates with lowest mean absolute error of 50.76 kN and Nash-Sutcliffe model efficiency coefficient of 0.984, compared to the most accurate predictions observed among the existing methods (modified Thornton method) with mean absolute error and Nash-Sutcliffe model efficiency coefficient of 227.22 kN and 0.872, respectively. Finally, a reliability analysis was performed according to the LRFD approach to identify the level of safety of predictions obtained by the CatBoost and XGBoost models. The CatBoost model exhibited remarkably low dispersion with a safety factor of 0.753 thereby providing an adequate level of safety for the design of gusset plates.
Design-focused Interpretable Machine Learning Models for Compressive Capacity Prediction of Gusset Plate Connections
https://orcid.org/0000-0001-8012-8925
https://orcid.org/0000-0002-2561-993X
Highlights An interpretable machine learning framework is developed to design gusset plate connection. A comprehensive dataset consisting of experimental tests and numerical simulation is assembled. Boosting algorithms (XGBoost, CATBoost) demonstrated very high accuracy in predicting the gusset plate capacity and outperformed existing empirical models. Feature importance as well as the global and local explanations of the developed models are explained employing SHAP. A reliability analysis is performed to identify the level of safety of predictions obtained by the CatBoost and XGBoost models.
Abstract Gusset plates can experience large compressive loads when used to connect bracing systems to beam-column joints in steel frames and truss bridges. The strength of a gusset plate is therefore vital to allow the bracing members to reach their ultimate capacity without a brittle failure of the gusset plate connection. Due to the large complexity of the stress distribution mechanism, simplified approaches commonly adopted for design can lead to inaccurate and highly conservative results. The present study aims at developing a simple yet efficient gusset plate compressive strength prediction model that does neither rely on crude assumptions nor require complicated calculations. A meticulously compiled database consisting of 68 experimental and 184 numerical datasets was deployed to aid in developing a high-performance machine learning (ML) model selected amongst 10 machine learning techniques studied. The superiority of the CatBoost and XGBoost models was confirmed with an accuracy of 95% and 96%, respectively significantly outperforming current design practices and widely used empirical formulas. SHapley Additive exPlanations (SHAP) and Taylor diagram were used to explain the feature importance and model performance. The developed interpretable ML models can instill trust in a model that explains its decisions, as opposed to other black box approaches. It was found that CatBoost was the most efficient in predicting the compressive strength of gusset plates with lowest mean absolute error of 50.76 kN and Nash-Sutcliffe model efficiency coefficient of 0.984, compared to the most accurate predictions observed among the existing methods (modified Thornton method) with mean absolute error and Nash-Sutcliffe model efficiency coefficient of 227.22 kN and 0.872, respectively. Finally, a reliability analysis was performed according to the LRFD approach to identify the level of safety of predictions obtained by the CatBoost and XGBoost models. The CatBoost model exhibited remarkably low dispersion with a safety factor of 0.753 thereby providing an adequate level of safety for the design of gusset plates.
Design-focused Interpretable Machine Learning Models for Compressive Capacity Prediction of Gusset Plate Connections
https://orcid.org/0000-0001-8012-8925
https://orcid.org/0000-0002-2561-993X
Highlights An interpretable machine learning framework is developed to design gusset plate connection. A comprehensive dataset consisting of experimental tests and numerical simulation is assembled. Boosting algorithms (XGBoost, CATBoost) demonstrated very high accuracy in predicting the gusset plate capacity and outperformed existing empirical models. Feature importance as well as the global and local explanations of the developed models are explained employing SHAP. A reliability analysis is performed to identify the level of safety of predictions obtained by the CatBoost and XGBoost models.
Abstract Gusset plates can experience large compressive loads when used to connect bracing systems to beam-column joints in steel frames and truss bridges. The strength of a gusset plate is therefore vital to allow the bracing members to reach their ultimate capacity without a brittle failure of the gusset plate connection. Due to the large complexity of the stress distribution mechanism, simplified approaches commonly adopted for design can lead to inaccurate and highly conservative results. The present study aims at developing a simple yet efficient gusset plate compressive strength prediction model that does neither rely on crude assumptions nor require complicated calculations. A meticulously compiled database consisting of 68 experimental and 184 numerical datasets was deployed to aid in developing a high-performance machine learning (ML) model selected amongst 10 machine learning techniques studied. The superiority of the CatBoost and XGBoost models was confirmed with an accuracy of 95% and 96%, respectively significantly outperforming current design practices and widely used empirical formulas. SHapley Additive exPlanations (SHAP) and Taylor diagram were used to explain the feature importance and model performance. The developed interpretable ML models can instill trust in a model that explains its decisions, as opposed to other black box approaches. It was found that CatBoost was the most efficient in predicting the compressive strength of gusset plates with lowest mean absolute error of 50.76 kN and Nash-Sutcliffe model efficiency coefficient of 0.984, compared to the most accurate predictions observed among the existing methods (modified Thornton method) with mean absolute error and Nash-Sutcliffe model efficiency coefficient of 227.22 kN and 0.872, respectively. Finally, a reliability analysis was performed according to the LRFD approach to identify the level of safety of predictions obtained by the CatBoost and XGBoost models. The CatBoost model exhibited remarkably low dispersion with a safety factor of 0.753 thereby providing an adequate level of safety for the design of gusset plates.
Design-focused Interpretable Machine Learning Models for Compressive Capacity Prediction of Gusset Plate Connections
Rahman, Jesika (Autor:in) / Billah, AHM Muntasir (Autor:in) / Arafin, Palisa (Autor:in) / Islam, Kamrul (Autor:in) / Nehdi, Moncef L. (Autor:in)
Engineering Structures ; 298
10.10.2023
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Gusset Plate Compression Capacity Prediction Using Ensemble Machine Learning Models
| Springer Verlag | 2023
Behavior of Gusset Plate Connections Under Compressive Monotonic and Cyclic Loadings
| British Library Conference Proceedings | 1993
Seismic design of braced frame gusset plate connections
| British Library Conference Proceedings | 2005