目的 采用近红外光谱分析技术建立干姜中多指标成分含量快速预测方法，对不同产地干姜进行快速无损的质量评价，提高干姜的质量控制水平。方法 建立同时测定干姜中6-姜酚、8-姜酚、10-姜酚和6-姜烯酚等4个主要活性成分的超高效液相色谱方法，并以其测定值为参比；采集不同产地干姜的近红外光谱，比较筛选出最优的光谱预处理方法，采用联合区间偏最小二乘法优选出最佳光谱区间，构建干姜各指标成分的最优偏最小二乘回归（partial least squares regression，PLSR）定量模型。结果建立的干姜中6-姜酚、8-姜酚、10-姜酚和6-姜烯酚最佳PLSR定量模型的校正决定系数（R_C²）分别为0.973、0.980、0.979和0.938，预测决定系数（R_P²）分别为0.926、0.920、0.883和0.781，4个定量模型的预测相对分析误差（residual predictive deviation,RPD）均大于2，表明建立的近红外光谱定量模型的预测值与测定值具有良好的线性关系，模型预测效果良好。结论 所建立近红外光谱定量模型，可以实现干姜中6-姜酚、8-姜酚、10-姜酚和6-姜烯酚等4个主要活性成分含量的快速预测，方法简便快捷，结果准确可靠，可为干姜质量的快速评价提供依据。

Objective To establish a rapid qualitative method for evaluating the content of multi-index components quickly and non-destructively by using near infrared spectroscopy (NIRs), and improve the quality control level of Zingiberis Rhizoma from different geographical origins.Methods An ultra performance liquid chromatography (UPLC) method was established to determine the contents of 6-gingerol, 8-gingerol, 10-gingerol and 6-shogaol in Zingiberis Rhizoma, which were adopted as the reference value. The NIRs of the samples from different geographical origins were collected. The different spectral pretreatment methods were compared and the best one was selected. Synergy interval-PLS was used to screen the characteristic spectral interval to obtain the best partial least squares regression (PLSR) model of each index component of Zingiberis Rhizoma.Results The coefficient of determination for calibration (R_C²) for the best PLSR models established for the quantitative determination of 6-gingerol, 8-gingerol, 10-gingerol, and 6-shogaol in Zingiberis Rhizoma was 0.973, 0.980, 0.979, and 0.938, respectively, while the coefficient of determination for prediction (R_P²) was 0.926, 0.920, 0.883, and 0.781, respectively. The values of residual predictive deviation (RPD) of the four final optimized PLSR models were greater than 2. The results suggested that the predicted values of NIR models and the measured values showed a good linear relation, indicating a great prediction ability of the models.Conclusion The established NIRs quantitative model could realize the rapid determination of the contents of 6-gingerol, 8-gingerol, 10-gingerol, and 6-gingerol in Zingiberis Rhizoma. This method is simple and fast, and the results are accurate and reliable, which can serve as a reference for rapid quality evaluation of Zingiberis Rhizoma.

R286

国家重点研发计划项目(2020YFC1712700);财政部和农业农村部:国家现代农业产业技术体系资助(CARS-21);江苏省“333高层次人才培养工程”;江苏省高校“青蓝工程”