基于时间序列的鄱阳湖水位分析与预测  

Analysis and Forecast of Poyang Lake's Water Level Base on Time Series

作　　者：张飞[1] 王建强[1,2,3] 罗寒[1] 

机构地区：[1]东华理工大学测绘工程学院,江西南昌330013 [2]流域生态与地理环境监测,国家测绘地理信息局重点实验室,江西南昌330013 [3]江西省数字国土重点实验室,江西南昌330013

出　　处：《测绘与空间地理信息》2016年第8期35-37,41,共4页Geomatics & Spatial Information Technology

基　　金：国家自然科学基金项目(41304020,41204003,41404026,41464001);东华理工大学博士科研启动基金项目(DHBK201114);江西省研究生创新专项资金资助项目(YC2015-S263)资助

摘　　要：以鄱阳湖外洲水文站为研究对象,主要介绍了水位与水流量的曲线模型以及自回归模型,选择短周期的水位和水流量数据,采用Matlab拟合最优的水位与水流量的曲线模型,剔除水流量对水位的影响,采用自回归模型并基于最小二乘原理对其剩余量进行拟合及预测。结果显示,曲线模型+自回归模型的拟合和预测效果明显优于单一的曲线模型和自回归模型,用后验差法检验显示预测精度等级为好,因此,该模型可对鄱阳湖外洲站水位进行短期预报。Take Waizhou hydrological station of Poyang Lake as research target,The paper mainly introduces the curve mode of water level and water flow and the autoregressive model,choosing a short period of water level and water flow data,fitting the best curve model of water level and water flow by using Matlab,excluding the impact of water flow on the water level,fitting and forecasting the remaining amount by using autoregressive model and the principle of least squares.The result shows curve model ＋ autoregressive model＇s effect of fitting and forecasting obviously better than single curve model and autoregressive model,showing that the prediction accuracy is good by using the after-test residue checking,so the model can be short-term forecast for Waizhou hydrological station of Poyang Lake.

关 键 词：鄱阳湖 水位 最小二乘原理 自回归模型 

分 类 号：P207[天文地球—测绘科学与技术]