针对地球自转变化复杂的时变特性，本项目采用高效的灰色神经网络组合模型来预报地球定向参数(EOP)。由于固体地球及环绕着它的流体圈层构成一个近似封闭的动力学系统，角动量守恒原理表明，大气或海洋角动量的任何变化都会影响固体地球的自转变化。将大气、海洋角动量时间序列引入到地球自转变化预报中，相当于增加物理约束条件。正是基于此，本项目着重研究和探索应用灰色神经网络组合模型，将大气和海洋角动量时间序列同时引入到地球自转变化预报中，改善地球定向参数的预报精度，本项目：（1）将灰色系统思想与神经网络有机地结合起来，构成灰色神经网络，发挥两者的优势，对提高预报效率具有重要意义；（2）将大气、海洋引入EOP预报，对提高EOP预报精度和稳定性具有重要的科学意义；（3）对于丰富EOP预报理论，提高深空探测器导航与追踪精度、维持坐标系统和时间系统，建立我国自主的高精度EOP预报方法和系统具有重要的科学和现实意义。

In view of the complex time-variable characteristics of the Earth’s variable rotation,this study employs the efficient Grey- General Regression Neural Network (G-GRNN) to predict the Earth Orientation Parameter (EOP). As the solid Earth and its surrounding fluid layers form an approximately close dynamic system, changes of atmospheric or oceanic angular momentum (OAM) will result in variations in the solid Earth’s rotation, according to the conservation law of angular momentum. When the atmospheric angular momentum (AAM) and OAM series are incorporated into the prediction of the Earth’s variable rotation, it will impose physical constraints to the prediction. This project focuses on incorporating the AAM and OAM series into the prediction of the Earth’s variable rotation to improve accuracies of the EOP predictions by G-GRNN model. This study will contribute to improve the efficiency, accuracy and stability of EOP prediction by fusing grey model and GRNN, and introducing of the AAM and OAM series. This will inrich the theory of EOP prediction, improve the accuracy of high-space navigation, tracking, and coordinate and time system maintainment, and benefit the construction of our own high accuracy EOP prediction method and system.