地球自转参数极移的预报方法研究

发布时间：2018-07-02 07:57

本文选题：极移预报 + 最小二乘外推　；参考：《山东科技大学》2017年硕士论文

【摘要】：高精度的极移参数具有重要的理论研究意义和实用价值。但是,测量数据需要通过复杂的解算过程,才能得到最终的极移参数,因此极移是不能实时获取的。为了满足极移的实时需要,对极移的预报的研究尤为必要。但是,由于极移的激发机制相对复杂,导致其预报结果并不是很理想。因此,极移的预报仍然需要进行进一步的深入研究。随着极移序列观测精度的提高,以及极移预报方法的优化,极移的预报精度正在逐步提高。但是关于极移预报的方法还有待进一步改进。基于此,本文做了如下工作:(1)为了进一步提高极移的短期预报精度,本文进一步研究了 ARMA模型族在极移预报中的应用,提出了用LS+ARMA对极移进行1-5天的短期预报,通过与其他方法的预报结果对比发现,本文精度均好于其他学者的预报结果,验证了该模型的可行性。(2)从数据平稳性出发,提出了基于双差分LS+ARMA模型的预报方法,用于极移1-100天跨度上的预报。并与直接用LS+ARMA模型进行预报的结果进行比较,结果显示,双差分LS+ARMA模型的预报结果在100天跨度上,在1-100天跨度上的预报结果均优于直接用LS+ARMA模型进行预报。(3)考虑到存在高频信号会对极移预报精度产生一定的影响,提出了采用小波分析将极移序列分解成一个低频信号与多个高频信号,通过不同的频率组合方案进行极移预报,来研究高频信号对极移预报精度的影响。从统计结果可以看出,高频信号会在一定程度上降低极移的短期预报精度,但是对极移的中长期预报精度的影响几乎可以忽略不计,总来看,低频信号与相对应高频信号的组合预报结果最优。此外,用该方法对极移进行1-365天跨度上的预报,并与IERS发布的预报结果进行对比,结果表明,该方法优于IERS的预报结果,验证了该方法的可行性与有效性。
[Abstract]:High precision polar motion parameters have important theoretical significance and practical value. However, the measured data can only be obtained by complex calculation process, so the pole shift can not be obtained in real time. In order to meet the real-time needs of the pole shift, it is necessary to study the prediction of the pole shift. However, due to the complexity of the excitation mechanism of the pole shift, the prediction results are not very satisfactory. Therefore, the prediction of the pole shift still needs to be further studied. With the improvement of the observation accuracy of the polar shift sequence and the optimization of the polar motion prediction method, the prediction accuracy of the pole shift is gradually improved. However, the method of pole shift prediction needs further improvement. Based on this, this paper has done the following work: (1) in order to further improve the accuracy of short-term polar motion prediction, this paper further studies the application of ARMA model family in polar shift prediction, and puts forward a short-term forecast of polar shift using LS ARMA for 1-5 days. By comparing the prediction results with other methods, it is found that the accuracy of this paper is better than that of other scholars, and the feasibility of the model is verified. (2) based on the data smoothness, a prediction method based on double difference LS ARMA model is proposed. Used for the prediction of the polar shift over the span of 1-100 days. The results are compared with the results obtained by using the LS ARMA model directly. The results show that the prediction results of the double difference LS ARMA model are on a 100-day span. The prediction results on 1-100 days span are better than those from LS ARMA model directly. (3) considering the existence of high frequency signals, the accuracy of polar motion prediction is affected to a certain extent. The polar shift sequence is decomposed into one low frequency signal and several high frequency signals by wavelet analysis. The influence of high frequency signal on the precision of pole shift prediction is studied by different frequency combination schemes. It can be seen from the statistical results that the high frequency signal will reduce the short-term prediction accuracy of the pole shift to some extent, but the effect on the long-term prediction accuracy of the pole shift can be almost ignored. The combination prediction results of low frequency signal and corresponding high frequency signal are optimal. In addition, the method is used to predict the polar shift over 1-365 days span, and the results are compared with the predicted results released by IERS. The results show that this method is superior to the prediction results of IERS, and the feasibility and effectiveness of the method are verified.
【学位授予单位】：山东科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P183.31;P228

【参考文献】