论文部分内容阅读
为了提高近红外光谱分析精度和速度,需要分别对近红外光谱扣除背景、降低噪音和优选变量等预处理,其过程稍嫌烦琐,本文提出用小波变换-偏最小二乘法(PLS)。该方法利用小波变换提取光谱变量,利用PLS将光谱变量和柴油性质进行关联,建立模型。利用该方法能够分析柴油的诸多性质,分析精度与传统处理方法(微分-PLS)基本一致。该方法具有预处理简单、优选参数少、建模变量少等特点,能够大大简化建模过程、提高建模和分析速度。
In order to improve the accuracy and speed of near infrared spectroscopy, pretreatment such as background subtraction, noise reduction and optimization variables needs to be performed separately. The process of this method is somewhat cumbersome. In this paper, wavelet transform-partial least squares (PLS) is proposed. The method uses wavelet transform to extract spectral variables, and uses PLS to correlate the spectral variables with the properties of diesel to establish a model. The method can be used to analyze many properties of diesel oil, the analysis accuracy is basically the same with the traditional treatment method (differential-PLS). This method has the advantages of simple preprocessing, less parameters optimization and less modeling variables, which can greatly simplify the modeling process and speed up the modeling and analysis.