基于PCA-BP神经网络的EDXRF分析测定地质样品中铁、钛元素含量的应用研究

Research on the Application of Principal Component Analysis and Improved BP Neural Network to the Determination of Fe and Ti Contents in Geological Samples

为实现地质样品中元素含量的准确预测,提出了基于主成分分析(PCA)的改进型BP神经网络模型。采用X荧光光谱法,对新疆西天山地质样品中Fe,Ti,V,Pb和Zn等元素进行测量,将得到的X荧光计数作为输入变量,应用该模型对未知地质样品中Fe和Ti元素进行定量预测。结果表明:主成分分析与改进型BP神经网络模型取得了较好的预测效果,预测结果与化学分析值的相对误差小于3%,为地质样品元素含量预测提供了一种新型有效的方法。

Aiming at forecasting elemental contents in geological samples accurately, a principal component analysis and im- proved BP （PCA-BP） neural network theory is proposed in the present work. The samples from west Tianshan were measured through X-ray fluorescence measurement method, and the X-Ray fluorescence counts of each element such as Fe, Ti, V, Pb, Zn, etc. were input to the PCA-BP neural network as input variables to forecast Fe and Ti contents in uncertified geological sam- ples quantitatively. The results show that the PCA-BP neural network can give an ideal result, and the relative error between the forecast data and chemical analysis data is less than 3 0%. This method provides a new and effective approach to forecasting ele- mental contents in geological samples.