Abstract

Data-based dam monitoring model can predict dam behavior and provide scientific basis for risk assessment and decision-making of dam engineering. Traditional dam safety monitoring models established by multiple regression, stepwise regression, gray theory, and other statistics-based methods have poor robustness and generalization. For this reason, an optimized support-vector machine (SVM) whose novelty lies in simple implementation, self-adaptive hyperparameter selection with aid of adaptive position particle swarm optimization (APPSO), and retention of influence factor combination is presented for constructing the dam safety monitoring model. A real-world concrete gravity dam engineering is used to verify the presented approach, and the results indicate the monitoring model built by APPSO–SVM outperforms benchmark models in terms of fitting and prediction accuracy.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call