摘要
水文预报能够为水资源的保护提供强有力的依据。本文将经验模式分解和动态递归神经网络相结合建立多步预测模型,以此来解决预测数据的非线性、精度低等问题。最后通过对比实验来说明,本文算法能够通过经验模式分解得到不同的模式分量,降低了原始数据之间的耦合性,提高了系统的稳定性。
Hydrological forecasting can provide a strong basis for the protection of water resources. In this paper, the empirical mode decomposition and dynamic recurrent neural network were combined to establish a multi-step prediction model, in order to solve the problem of nonlinear prediction data and low accuracy. Finally, through the comparison experiments, we could show that the algorithm could decompose the different model components through empirical mode, reduce the coupling between the original data and improve the stability of the system.
出处
《舰船科学技术》
北大核心
2017年第2X期136-138,共3页
Ship Science and Technology
关键词
多步预测
经验模式分解
动态递归神经网络
multi-step prediction
empirical mode decomposition
dynamic recurrent neural network
