SSA-Optimized Ensemble Learning Models for Accurate and Interpretable Crest Settlement Prediction of Concrete-Faced Rockfill Dams

Xiaoyuan Li¹, Ming Xu¹, Shibin Yao¹, Su Wang^2,*, Jian Zhou^1,*
1 School of Resources and Safety Engineering, Central South University, Changsha, China
2 Kunming Prospecting Design Institute of China Nonferrous Metals Industry Co., Ltd., Kunming, China
* Corresponding Author: Su Wang. Email: email ; Jian Zhou. Email: email
(This article belongs to the Special Issue: Computational Intelligent Systems for Solving Complex Engineering Problems: Principles and Applications-III)

Computer Modeling in Engineering & Sciences https://doi.org/10.32604/cmes.2026.086451

Received 30 May 2026; Accepted 31 August 2026; Published online 17 September 2026

Download PDF

Abstract

Accurate prediction of crest settlement in Concrete-Faced Rockfill Dam (CFRD) is of great significance for safety during its construction and operational phases. In this study, the Squirrel Search Algorithm (SSA) was used to optimize Random Forest (RF), EXtreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LGBM) to improve model performance. The original dataset, containing 74 cases, was augmented to train and test the models. The final results showed that among all the developed models, the XGBoost model optimized by SSA with a population size of 25 achieved the best performance, with an R² of 0.936, RMSE of 0.0210, MAE of 0.0152, VAF of 93.62%, and an A-20 index of 0.830 on the test set. Furthermore, several different model interpretation techniques were employed to analyze the influence of different input variables on the predictions. The analysis results indicated that dam height (H) is the most influential parameter on crest settlement. In conclusion, the predictive model developed in this study effectively predicts crest settlement and demonstrates interpretability.