基于经验模式分解卡尔曼滤波的径流预报校正

摘要

针对水文径流预测过程中不可避免的误差，提出基于经验模式分解的卡尔曼滤波校正方法，采用经验模式分解理论将径流序列转变为平稳过程，并结合预报过程误差规律及不同物理背景，运用基于Sigma点的卡尔曼滤波校正的动态系统分析技术，对单个水文模型的输出资料进行同化操作，对状态方程进行多变量最优控制，实时修正径流预报误差。实时校正结果表明，基于经验模式分解的卡尔曼滤波校正方法能一定程度上提高径流预报精度。%For the inevitable error existing in hydrological runoff prediction, a Kalman filter correction method based on empirical mode decomposition is presented, in which, the empirical mode decomposition is used to transform runoff into stationary process, and then combined with prediction error rule and different physical process background, the corrected dynamic system analysis technique based on Sigma point Kalman filtering is used to assimilate the output data of a single hydrological model and achieve the optimal control of multi-variable state equation to real-time correct runoff forecasting error. The real-time correction results of model show that the Kalman filter correction method based on empirical mode decomposition can improve runoff forecast accuracy.