基于神经网络的恒星大气参数自动测量

英文篇名：Automatic prediction of stellar atmospheric parameters with neural network
中文刊名：辽宁科技大学学报
英文刊名：Journal of University of Science and Technology Liaoning
作者：潘亚春 ; 屠良平
英文作者：PAN Ya-chun1 ; TU Liang-ping2(1.School of Materials Science and Engineering ; University of Science and Technology Liaoning ; Anshan 114051 ; China ; 2.School of Science ; University of Science and Technology Liaoning ; Anshan 114051 ; China)
中文关键词：恒星大气参数 ; 人工神经网络 ; 数据降维
英文关键词：stellar atmospheric parameters ; artificial neural network ; data dimension reduction
出版日期：2009-02-15
机构：辽宁科技大学材料科学与工程学院;辽宁科技大学理学院;
年：2009
期：01
出版单位：辽宁科技大学学报

摘要

提出了一种基于神经网络的恒星光谱大气参数的自动测量方法,该方法能够根据恒星的光谱自动得到恒星的三个重要参数:有效温度Teff,表面重力logg和金属丰度[Fe/H]的估计值。首先对实测恒星光谱进行预处理,包括小波去噪,光谱波长统一化;然后通过对训练样本进行PCA分析获得特征空间变换矩阵进行数据降维;最后通过训练好的神经网络得到参数的估计值。实验结果分析表明,该方法比其它估计方法如非参数估计、支持向量回归和偏最小二乘回归具有更高的测量精度。
An automatic method to predict stellar atmospheric parameters using artificial neural network(ANN) was presented,which can achieve the prediction of stellar effective temperature Teff,surface gravity log g and metallic abundance.The method is composed of three functional units,namely stellar spectra preprocessing(include stellar spectra denoise and wavelength normalization),data reduction(using principal components analysis) and parameters estimation using BP network.The experimental results show that the method is better than other methods such as non-parameter estimation,SVM regression and PLS regression.

引文

[1]LAMOST项目[DB/OL].http://www.lamost.org/.
    [2]李宗伟,肖兴华.天体物理学[M].北京:高等教育出版社,2000:1.
    [3]SOUBIRAN D,KATZ D,CAYREL R.On-line determination of stellar atmospheric parametersTeff,logg,[Fe/H]fromELODIE echelle spectra I[J].Astronomy&Astrophysics,1998,(338):151-160.
    [4]KATZ D,SOUBIRAN C,CAYREL R,et al.On-line determination of stellar atmospheric parametersTeff,logg,[Fe/H]from ELODIE echelle spectra II[J].Astronomy and Astrophysics Supplement,1998,(133):221-226.
    [5]FUENTES O,GULATI R K.Prediction of stellar atmospheric parameters from spectra,spectral indices and spectral linesusing machine learning[C].RevmexAA(serie de conferencias)2001,(10):209-212.
    [6]BAILER-JONES C A L.Determination of stellar parameters with GAIA[J].Astrophysics and Space Science,2002,(280):21-29.
    [7]BAILER-JONES C.Stellar parameters from very low resolution spectra and medium band filtersTeff,loggand[M/H]us-ing neural networks[J].Astronomy and Astrophysics,2000,(357):197-205.
    [8]ADELMAN M,JENNIFER K,AGUEROS M A,et al.The sixth data release of the sloan digital sky survey[J].The Astro-physical Journal Supplement Series,2008,175(2):297-313.
    [9]王碧泉,陈祖荫.模式识别理论、方法及应用[M].北京:地震出版社,1989:39-62.
    [10]WANG Y,WITTEN I H.Pace regression.Technical Report99/12[R].Department of Computer Science,University ofWaikato,1999.