Welcome to Acta Armamentarii ! Today is

Acta Armamentarii ›› 2025, Vol. 46 ›› Issue (11): 250019-.doi: 10.12382/bgxb.2025.0019

Previous Articles     Next Articles

Prediction Model of Concrete Penetration Depth Based on Unbiased Feature Selection

ZHANG Xiuli1,2, ZHANG Jie1,2,*(), QI Xiaopeng1,2, WANG Zhiyong1,2, LIU Zhifang1,2, WANG Zhihua1,2   

  1. 1 Institute of Applied Mechanics, College of Aeronautics and Astronautics, Taiyuan University of Technology, Taiyuan 030024, Shanxi, China
    2 Shanxi Key Laboratory of Material Strength and Structural Impact, Taiyuan 030024, Shanxi, China
  • Received:2025-01-06 Online:2025-11-27
  • Contact: ZHANG Jie

Abstract:

To address the challenges of insufficient data,complex feature selection,and limited applicability of traditional empirical formulas in the prediction of concrete penetration depth,this paper proposes a concrete penetration depth prediction model based on deep learning and optimization strategies.Initially,218 sets of concrete penetration test data are collected,and the size and diversity of dataset are expanded through the integration of multi-stage empirical formulas.To enhance the accuracy of feature selection,an XGBoost (eXtreme Gradient Boosting) model combined with SHAP (SHapley Additive exPlanations) values is employed for feature screening and dimensionality reduction,thereby identifying the key features that are correlated with penetration depth.A multi-layer perceptron (MLP) model for penetration depth prediction is constructed based on these selected key features.Further,Bayesian optimization method is applied to fine-tune the hyperparameters of the MLP model,and its generalization capability is evaluated through K-fold cross-validation.The results demonstrate that the proposed model performs better than the traditional empirical formula under various conditions,its prediction accuracy and stability are significantly improved,and it shows good generalization ability and adaptability.In addition,compared with the traditional empirical formulas,the proposed model also shows better prediction effect in the applicable range of the empirical formulas.

Key words: concrete, penetration depth prediction, unbiased feature selection, Bayesian optimization

CLC Number: