宽度贝叶斯学习模型在抽水蓄能电站滑坡灾害预测的应用研究

In pumped-storage power station projects,accurate landslide prediction is essential for ensuring engineering safety and environmental protection.This study introduces the Broad Bayesian Learning(BBL)Model,designed to enhance the effectiveness of the Broad Learning System(BLS)on smaller datasets typical in such projects.The research data were sourced from feasibility and pre-feasibility study reports of a pumped-storage power station in Zhaoqing,Guangdong Province.This study primarily involved applying the BBL Model to specific engineering data and assessing its effectiveness and adaptability in real-world scenarios.The results indicate that the model delivers superior performance on small-scale datasets,maintaining high predictive accuracy even with limited data availability.Furthermore,the Bayesian algorithm within the model effectively quantifies uncertainties,thereby increasing its reliability for practical applications.This research extends the application scope of the BLS,offering a novel tool for landslide disaster risk assessment and prediction in small-scale engineering projects like pumped-storage power stations.These findings showcase the potential of advanced algorithms for effective disaster prediction and management in engineering design,particularly under resource-limited conditions,thus enhancing engineering safety.This study is expected to provide new theoretical and methodological contributions to the fields of geological disaster prevention and engineering safety.