正丁醇具有互溶性好、 挥发性低、 价格低廉以及腐蚀性低等优势, 被认为是理想的柴油添加物。 柴油中正丁醇的精准定量分析对其品质快速评价与市场监督具有重要科学意义与实用价值。 提出了一种基于拉曼（Raman）光谱结合偏最小二乘（PLS）的柴油中正丁醇快速定量分析方法。 首先, 采集了40个柴油样品的Raman光谱, 并考察了不同预处理方法［一阶导数、 二阶导数、 多元散射校正、 标准正态变换、 归一化（Normalization）和小波变换］对PLS校正模型预测性能的影响; 其次, 采用变量重要性投影（VIP）对Normalization方法预处理后的光谱数据进行特征变量提取, 并采用五折交叉验证优化VIP的阈值; 最后, 基于最优的光谱预处理方法、 输入变量和模型参数, 构建PLS校正模型对柴油中正丁醇含量进行快速定量分析, 结果与基于原始光谱（RAW）和Normalization光谱的PLS校正模型的预测性能进行对比。 结果表明: Normalization-VIP-PLS校正模型展现出优异的预测能力（R2CV和RMSECV为0.998 4和0.236 2%: R2P和RMSEP为0.998 7和0.208 4%; RSD为0.035 5）。 成功建立了一种Raman光谱结合PLS算法的柴油中正丁醇快速定量分析方法, 该方法具有快速、 准确、 便捷等优势, 可为柴油以及其他燃料中添加物检测及其品质分析提供新思路与新方法。

N-butanol is considered an ideal diesel additive because of its good solubility, low volatility, low price and corrosiveness. The accurate quantitative analysis of n-butanol in diesel has important scientific significance and practical value for its quality evaluation and market supervision. This paper proposes a rapid quantitative analysis method for n-butanol in diesel based on Raman spectroscopy combined with partial least squares (PLS). Firstly, Raman spectra of 40 diesel samples were collected, and the effects of different pretreatment methods (first derivative, second derivative, multivariate scattering correction, standard normal transform, Normalization and wavelet transform) on the prediction performance of the PLS calibration model were investigated. Secondly, variable importance in projection (VIP) is used to extract characteristic variables from the spectral data preprocessed by the Normalization method, and the threshold of VIP is optimized by five-fold cross-validation. Finally, based on the optimal spectral pretreatment method, input variables and model parameters, a PLS calibration model was built to analyze the content of n-butanol in diesel quantitatively. The prediction performance was compared with the RAW-PLS and Normalization-PLS models. The results show that the Normalization-VIP-PLS calibration model has excellent predictive performance (R2CV and RMSECV are 0.998 4 and 0.236 2%, R2P and RMSEP are 0.998 7 and 0.208 4%; RSD 0.035 5). Therefore, this paper successfully established a rapid quantitative analysis method of n-butanol in a diesel by Raman spectroscopy combined with the PLS algorithm. This method has the advantages of being fast, accurate and convenient and it can provide new ideas and methods for the detection and quality analysis of diesel and other fuel additives.