西江梧州站基于支持向量机的水位预测模型研究  

Study on Water Level Prediction Model Based on Support Vector Machine in Wuzhou Station of Xijiang River

作　　者：朱颖洁 ZHU Yingjie(Wuzhou Hydrological Center,Wuzhou 543002,China)

机构地区：[1]梧州水文中心,广西梧州543002

出　　处：《广东水利水电》2022年第11期39-42,共4页Guangdong Water Resources and Hydropower

基　　金：广西自然科学基金项目(编号:桂科基0991026);广西重点实验室科研项目(编号:桂科能0701K019);广西水利厅科技项目(编号:201618)。

摘　　要：利用西江梧州站的水位历史观测数据,基于支持向量机对西江梧州站的水位预测模型进行深入研究,以总体方差最小为原则优选分析各模型变量,基于逐日平均历史数据的变化趋势,对2019年的日平均水位变化进行预测,并与自回归模型进行模拟效果对比分析。研究表明:支持向量机水位预测模型预测效果比自回归模型好;支持向量机水位预测模型预测精度较高,对于提高航道水位预测精度具有一定的参考价值。Water level data is an important reference for waterway management and ship navigation.Water level prediction plays an important guiding role in flood control,ship route planning,navigation safety and navigation efficiency improvement.In this paper,based on the historical observation data of the water level of Xijiang Wuzhou station,the water level prediction model of Xijiang Wuzhou station is studied based on support vector machine.The model variables are optimized and analyzed based on the principle of minimum overall variance.Based on the trend of daily average historical data,the daily average water level change in 2019 is predicted,and the simulation effect is compared with the autoregressive model.The research shows that the prediction results of support vector machine water level prediction model are better than those of autoregressive model.The prediction accuracy of support vector machine water level prediction model is high,which has certain reference value for improving the prediction accuracy of waterway water level.

关 键 词：支持向量机 水位预测 梧州 西江 

分 类 号：P338[天文地球—水文科学]