基于智能机器学习的水电站生产安全运行风险识别研究

打开文本图片集

中图分类号：TV737；TP391 文献标志码：A

文章编号：1001-5922（2025）09-0180-04

Research on risk identification of safe operation of hydropower station based on intelligent machine learning

GE Jia，TIANRuochao，WANG Qian （Guoneng Daduhe Pillow Head Dam Power Generation Co.，Ltd.，Leshan 614799，Sichuan China）

Abstract：Traditional risk identification methodsoftenrelyon manual judgment andexperience，which have problems such aslow identification eficiencyand incomplete information acquisition，and therisk identification method based on intellgent machine learning can effectively ensure the safe operation of hydropower stations.A production safetyrisk identification model forhydropower station based on inteligent machine learning was established，and the key parameters and prediction accuracy of the model were determined.The results showed that the penalty factor and kermel parameter of the LSSVM model were 328.41254521 and1.10234521 after the PSO algorithm optimization.For15 groups of training samples，the accuracy of the PSO-LSSVM production security recognition model and the CNN production security recognition model was 100% ，and the accuracy of the KNN production security recognition model was 93.3% . For 10 sets of test samples，the prediction accuracy of the PSO -LSSVM production security identification model was 90% ，while the prediction accuracy of the CNN production security identification model and the KNN production security identification model was 80% .For the application verification samples，the prediction accuracy of the PSO - LSSVM production security identification model was 100% ，while the prediction accuracy of the CNN production security identification model and the KNN production security identification model was 80% Keywords：machinelearning;inteligent identification;hydroelectricpowerplants;productionsafety;algorithms

水电站是我国重要的清洁能源开发利用工程之一，担负着电能供给和水资源调节等重要任务[1-3]。（剩余4989字）