Journal of Hydroelectric Engineering
Online:
Published:
Abstract: Dam failure will cause a large number of life and property losses and huge environmental damage. Accurate and fast prediction of the risk level of earth-rock dams is of great significance for controlling the dam failure hazard of earth and rock dams. In this paper, K-Nearest Neighbor (KNN) algorithm is used to fill a large amount of missing data in the database, and Genetic Algorithm (GA) was introduced to optimize the hyperparameters of Light Gradient Boosting Machine (LightGBM), and a fast prediction model of earth-rock dam risk grade based on GA-LightGBM was established. Other evaluation indexes such as ROC (Receiver Operating Characteristic) curve and AUC (Area Under the Curve) value are used to verify the model accuracy, and it is compared with the traditional machine learning model. The study shows that the model has a high accuracy of 89.95%. The AUC value of the model is 0.977, which indicates that the model is better than the traditional prediction model in terms of applicability and prediction accuracy. The model has been also analyzed for global influencing factors and case studies using Shapley Additive Explanations analysis(SHAP), and the results showed that the frequency of inspection is one of the most important influencing factors leading to the risk of earth-rock dams.
LI Yanlong, ZHANG Yuchun, WANGg Ting, YIN Qiaogang, LIU Yunhe. Study on intelligent prediction analysis and model optimization of earth-rock dams risk level[J].Journal of Hydroelectric Engineering, 0, (): 0-.
0 / / Recommend
Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks
URL: http://www.slfdxb.cn/EN/
Cited