AI-Driven prediction of the impermeable boundary in karst rock mass for optimized anti-seepage curtain design

Lei Zhang¹Zhihua Tan^2*Ruilang Cao³Jing Wang²Yongchuan Zhao²Hui Tian²Bingxu Wang²Chunlei Peng⁴Xiaoqian Huang⁴

• ¹Yunnan Institute of Water & Hydropower Engineering Investigation and Design Co., Ltd., Kunming, China
• ²Yunnan Institute of Water & Hydropower Engineering Investigation, Design and Research, Kunming, China
• ³China Institute of Water Resources and Hydropower Research, Beijing, China
• ⁴CCC HongYu Water Conservancy Engineering Co., LTD., Changsha, China

Accurately determining the bottom boundary of anti-seepage curtains is critical for ensuring the integrity and performance of this key engineered composite structure in karst reservoirs. This study leverages artificial intelligence (AI) to address this materials design challenge. We developed hybrid models by integrating a Genetic Algorithm (GA) with Backpropagation (BP), Support Vector Machine (SVM), and Extreme Learning Machine (ELM) algorithms. These models were trained and validated using a comprehensive dataset from the Dehou Reservoir, incorporating critical material and hydrogeological properties of the karst rock mass. The results demonstrated that GA optimization significantly enhanced predictive performance. The GA-BP model achieved superior accuracy (R²=0.98, MSE=7.58). Furthermore, from an engineering safety perspective, the GA-SVM model provided the most reliable recommendations, frequently yielding conservative depth estimates. A comparative analysis with Random Forest, eXtreme Gradient Boosting and Light Gradient Boosting Machine validated the competitive advantage of our proposed models. This research underscores the potential of AI-driven approaches for the performance prediction and rational design of engineered geomaterial systems, offering a powerful tool for infrastructure projects in complex geological settings.