Abstract:The estimation of the maximum displacements of RC beams subjected to impact loads is of pivotal importance to define demand models to be employed in structural design. The dynamic behavior is quite complex and the maximum displacements can be obtained through experimental tests or high-fidelity FEM models, while simplified analytical models accounting for few degrees of freedom have generally low accuracy. In this context, this paper presents a study on the training and interpretation of a machine-learning model that strategically combines several algorithms for the said purpose. To train the model, a comprehensive database of 342 samples of rectangular simply supported beams under rigid impact loads is adopted. Seven ML regression algorithms are adopted in this study to develop a predictive model: (1) Support Vector Regression (SV R), (2) Gaussian Process Regression (GPR), (3) Random Forest Regressor (RFR), (4) XGBoost, (5) CatBoost, (6) LightGBM, and (7) Artificial Neural Networks (ANN). Finally, a stacking model combining SV R, GPR, RFR, XGBoost, CatBoost and ANN is built to improve accuracy, wherein 75% and 25% of the data are used for training and validation, respectively, in combination with the 5 cross-validation technique. This effort resulted in approximately 90% validation accuracy exceeding current mechanics-based/semi-empirical models. Finally, the SHapley Additive exPlanations (SHAP) algorithm is used to estimate the relative importance of the input features; the most influential factors are impact velocity and impactor mass which are representing the initial kinetic energy. Moreover, a full parametric analysis of the proposed model is provided and the influence of each input is discussed. The model is finally applied to perform a structural check for three damage levels. The proposed model can be used in practical applications where a reliable estimation of the maximum displacements of RC beams under impact loading is required.

Research on Particle Accelerator Control Network Deformation Prediction Algorithm with Machine Learning

Deformation Prediction Model of Large-Span Prestressed Structure for Health Monitoring Based on Robust Gaussian Process Regression

Research on fusion and deformation analysis of decade' HLS-II control network surveying data

Machine learning for online control of particle accelerators

Rapid Prediction Method for Nonlinear Expansion Process of Medical Vascular Stent

Interpretable Machine-Learning Models for Maximum Displacements of RC Beams under Impact Loading Predictions

Prediction and Compensation of Force-Induced Deformation for a Dual-Machine-based Riveting System Using FEM and Neural Network

High-precision and High-Efficiency Measurement Method of Accelerator Tunnel Control Network Based on Total Station Angle Observation

Initial Application of Machine Learning for Beam Parameter Optimization at the Hefei Light Source II

Beam data modeling of linear accelerators (linacs) through machine learning and its potential applications in fast and robust linac commissioning and quality assurance

CNN-GRU Network-based Force Prediction Approach for Variable Working Condition Milling Clamping Points of Deformable Parts

Research on Methods to Suppress the Error Accumulation in the Elevation Direction for Accelerator Tunnel Control Network

A robust smoothing analysis strategy for improving particle accelerator alignment and installation accuracy and efficiency based on HALF design

Research on Datumn Transformation Model in Particle Accelerator Component Installation

Research on the Optimization Methods of Common Points Layout for Survey and Alignment of Particle Accelerators

Using Machine Learning Methods to Analyze Data Related to CSNS RCS Beam Transmission Efficiency

Data-driven Deformation Prediction and Control for Existing Tunnels below Shield Tunneling

Predictive Control Modeling of ADS’s MEBT Using BPNN to Reduce the Impact of Noise on the Control System

Strain distribution prediction in UHPC beams using deep learning model

A Data-Driven Method for Predicting Deformation of Machined Parts Using Sparse Monitored Deformation Data

Prediction on Geometrical Characteristics of Laser Energy Deposition Based on Regression Equation and Neural Network