本文關(guān)鍵詞：神經(jīng)網(wǎng)絡(luò)在日長變化預(yù)報(bào)中的應(yīng)用研究　出處：《中南大學(xué)》2011年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 日長變化預(yù)報(bào) 人工神經(jīng)網(wǎng)絡(luò) 非線性 GA 廣義回歸

【摘要】：地球自轉(zhuǎn)參數(shù)預(yù)報(bào)對于天文學(xué)和測地學(xué)的理論研究和實(shí)際應(yīng)用具有重要意義。在地球自轉(zhuǎn)參數(shù)的預(yù)報(bào)中,日長變化的預(yù)報(bào)是難點(diǎn)。日長變化包含了復(fù)雜的非線性因素,已有的科學(xué)研究包括對它的線性預(yù)報(bào)和非線性預(yù)報(bào)。本文在對已有研究進(jìn)行學(xué)習(xí)分析的情況下,提出使用非線性神經(jīng)網(wǎng)絡(luò)(包括遺傳算法(GA)優(yōu)化的BP神經(jīng)網(wǎng)絡(luò)、廣義回歸神經(jīng)網(wǎng)絡(luò))對日長變化進(jìn)行預(yù)報(bào),并與已有研究成果進(jìn)行對比分析,得到一些有益的成果,為日長變化的研究增加了新的方法。 本文研究的主要內(nèi)容包括： (1)分析了BP神經(jīng)網(wǎng)絡(luò)用于日長變化的基本方法及其存在的不足之處,提出使用遺傳算法優(yōu)化BP神經(jīng)網(wǎng)絡(luò)的初始權(quán)值和閾值,防止網(wǎng)絡(luò)陷入局部極小,將遺傳算法優(yōu)化的神經(jīng)網(wǎng)絡(luò)用于日長變化預(yù)報(bào),并將預(yù)報(bào)結(jié)果與BP神經(jīng)網(wǎng)絡(luò)預(yù)報(bào)的結(jié)果進(jìn)行比較。 (2)由于GA優(yōu)化的神經(jīng)網(wǎng)絡(luò)對日長變化的預(yù)報(bào)需要不斷循環(huán)計(jì)算以得到最優(yōu)的權(quán)值和閾值,對于冗長天文數(shù)據(jù)的預(yù)報(bào)無疑增加了預(yù)報(bào)時間,這給獲取實(shí)時快速的日長變化數(shù)據(jù)帶來了困難。文中嘗試使用更為簡潔高效的廣義回歸神經(jīng)網(wǎng)絡(luò)(GRNN)對日長變化進(jìn)行預(yù)報(bào),這種網(wǎng)絡(luò)模型不需要循環(huán)迭代,是一種局部尋優(yōu)算法,不會陷入局部極小,算法容易實(shí)現(xiàn),并將預(yù)報(bào)結(jié)果與Schuh(2002)及EOP PCC(2010)的預(yù)報(bào)結(jié)果進(jìn)行比較。實(shí)驗(yàn)結(jié)果表明,將廣義回歸神經(jīng)網(wǎng)絡(luò)用于日長變化的預(yù)報(bào)是切實(shí)可行的。 (3)傳統(tǒng)的日長變化預(yù)報(bào)在選取樣本數(shù)據(jù)時,多為按跨度i(i為間隔,取1、2、3…n)取值,而神經(jīng)網(wǎng)絡(luò)模擬的是相關(guān)事物之間的相關(guān)性,相關(guān)性越大,神經(jīng)網(wǎng)絡(luò)越能夠獲得足夠的先驗(yàn)信息,從而得到更優(yōu)的輸出結(jié)果。對于日長變化數(shù)據(jù),距離越近的數(shù)據(jù)之間的相關(guān)性越大,而輸入樣本采用按跨度i輸入的方式勢必忽略一些相近數(shù)據(jù)之間的相關(guān)性,損失重要的先驗(yàn)信息。所以文中提出輸入樣本按連續(xù)輸入的方式選取,并與輸入方式按跨度i的方式選取進(jìn)行日長變化預(yù)報(bào)的結(jié)果進(jìn)行比較。結(jié)果表明,樣本采用按跨度輸入的方式在超短期預(yù)報(bào)中預(yù)報(bào)精度較高,樣本采用連續(xù)輸入的方式在短期和中期預(yù)報(bào)中預(yù)報(bào)精度較高。
[Abstract]:The earth rotation parameter prediction has important significance for theoretical research and practical application of astronomy and geodesy. In the prediction of earth rotation parameters, length of day forecast is difficult. LOD contains complex nonlinear factors, including the scientific research on its linear prediction and nonlinear prediction in this paper. Analysis on the existing research situation, put forward the use of nonlinear neural network (including genetic algorithm (GA) optimized BP neural network, generalized regression neural network) to predict the length of day, and compared with the existing research results, get some useful results on variation of length of day adds new methods.
The main contents of this paper are as follows:
(1) and the shortcomings of the BP neural network method for the analysis of the change in the length of the day, we use genetic algorithm to optimize BP neural network initial weights and thresholds, to prevent the network from falling into local minimum, the neural network optimized by genetic algorithm for the length of day forecast, and compare the prediction results with BP neural the network prediction results.
(2) the optimization of GA neural networks to forecast the variation of length of day loop calculations are needed to obtain the optimal weights and thresholds for lengthy astronomical data forecast will increase the forecast time, this to obtain real-time data changes in length of day is difficult. This paper try to use a more generalized regression neural network simple and efficient (GRNN) to predict the length of day, this network model does not need iteration, is a local optimization algorithm, not into the local minimum, the algorithm is easy to realize, and the prediction results with Schuh (2002) and EOP PCC (2010) compared the prediction results. The experimental results show that the generalized regression neural network is used to forecast the length of day is feasible.
(3) when selecting sample data, the traditional day length variation forecast is mostly based on the span of I (I interval, 1,2,3... The n value, and the neural network simulation) is the correlation between things, the greater the correlation, neural network can obtain sufficient prior information, in order to obtain the output results better. For the change in the length of the day, the closer the relationship between the data and the greater input samples according to the span of the I input the way will ignore some correlations between similar data, loss of priori information. So in this paper the input samples according to the continuous input mode selection, and the input by way of I span was selected for the prediction of LOD were compared. The results showed that the samples according to the input span high prediction accuracy in ultra in the short-term forecasting, the sample with the continuous input mode in the short and medium term prediction accuracy is high.

【學(xué)位授予單位】：中南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2011
【分類號】：TP183;P183.31

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 周永宏,鄭大偉,廖新浩;日長變化、大氣角動量和ENSO:1997～1998厄爾尼諾和1998～1999拉尼娜信號[J];測繪學(xué)報(bào);2001年04期

2 馬騰;王耀強(qiáng);花向紅;;遺傳神經(jīng)網(wǎng)絡(luò)在GPS高程轉(zhuǎn)換中的應(yīng)用[J];測繪信息與工程;2008年03期

3 李杰;王科;王航;;基于廣義回歸神經(jīng)網(wǎng)絡(luò)的公路貨運(yùn)量預(yù)測方法研究[J];交通與計(jì)算機(jī);2007年03期

4 ;Real-time rapid prediction of variations of Earth's rotational rate[J];Chinese Science Bulletin;2008年07期

5 鄭大偉;陳兆國;;地球自轉(zhuǎn)參數(shù)預(yù)測[J];中國科學(xué)院上海天文臺年刊;1982年00期

6 顧暉;鄭大偉;;地球自轉(zhuǎn)、大氣角動量和太陽活動的30-60天波動特征[J];中國科學(xué)院上海天文臺年刊;1992年00期

7 楊志根;;地球自轉(zhuǎn)日長變化中某些短周期項(xiàng)的可能成因[J];中國科學(xué)院上海天文臺年刊;1992年00期

8 王琪潔;廖德春;周永宏;廖新浩;;日長變化預(yù)報(bào)中BP神經(jīng)網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)的選擇[J];中國科學(xué)院上海天文臺年刊;2007年00期

9 廖德春;王琪潔;周永宏;廖新浩;;應(yīng)用神經(jīng)網(wǎng)絡(luò)技術(shù)預(yù)報(bào)海表溫度變化異常[J];中國科學(xué)院上海天文臺年刊;2008年00期

10 鄭大偉;時間序列分析與地球自轉(zhuǎn)[J];天文學(xué)進(jìn)展;1989年02期

相關(guān)博士學(xué)位論文 前1條

1 劉清;人工神經(jīng)網(wǎng)絡(luò)算法的改進(jìn)及其在水利測量技術(shù)中的應(yīng)用[D];河海大學(xué);2005年

相關(guān)碩士學(xué)位論文 前1條

1 陳玉紅;RBF網(wǎng)絡(luò)在時間序列預(yù)測中的應(yīng)用研究[D];哈爾濱工程大學(xué);2009年

，

本文編號：1422417