基于集合卡尔曼滤波的入库流量异常波动改进  

Correction of Abnormal Fluctuations in Reservoir Inflow Data based on Ensemble Kalman Filter

作　　者：莫和建 朱喜 张津 张俨 刘杨洋 MO Hejian;ZHU Xi;ZHANG Jin;ZHANG Yan;LIU Yangyang(Remote Control Center for Hydropower Stations,Guizhou Wujiang Hydropower Development Co.,Ltd.,Guiyang 550002,China;NARI Technology Co.,Ltd.,Nanjing 210000,China;Power Dispatching Control Center,Guizhou Power Grid Co.,Ltd.,Guiyang 550002,China)

机构地区：[1]贵州乌江水电开发有限责任公司水电站远程集控中心,贵州贵阳550002 [2]国电南瑞科技股份有限公司,江苏南京210000 [3]贵州电网有限责任公司电力调度控制中心,贵州贵阳550002

出　　处：《水电与新能源》2025年第2期42-47,共6页Hydropower and New Energy

摘　　要：为了平滑反推的小时入库流量的异常锯齿状波动,基于水库水量平衡原理,提出了物理机制驱动的集合卡尔曼滤波(EnKF)方法。以乌江渡水库为例,验证了该方法对典型入库流量过程中异常波动的修正效果。研究结果表明:EnKF方法处理后的入库流量序列在平滑度和上下库水量约束指数方面均优于三点滑动平均法和五点三次滑动平均法。To smooth out the abnormal zigzag fluctuations in reservoir hourly inflow data,a physical mechanism driven ensemble Kalman filter(EnKF)method is proposed based on the reservoir water balance principle.Taking the Wujiangdu Reservoir as an example,the correction effect of the proposed method on abnormal fluctuations in typical inflow processes is verified.The results show that the EnKF method outperforms the three-point moving average method and the five-point three-order moving average method in terms of the smoothness and the upper and lower reservoir water storage constraint index of the corrected inflow sequence data.

关 键 词：入库流量 梯级水库 集合卡尔曼滤波 数据平滑 乌江流域 

分 类 号：P33[天文地球—水文科学]