Abstract The test is to control quality of Gentiana straminea Maxim. by near infrared spectroscopy. Samples were collected throughout Qinghai-Tibet Plateau, in which the bioactive ingredients were gentiopicroside and loganic acid. Sample set was divided using concentration gradient method. Modeling optimization was processed with a designed orthogonal test L9 (33). Pretreatment of spectra was multiplicative scatter correction + second derivative, and partial least square was used to establish model. The model showed high R and low root mean square error. Moreover, residual predictive deviation was 3.14. The present quantitative model can fast and accurately determine the quality of G. straminea.

中文翻译：

---

一种快速评价中药龙胆质量的新方法。使用近红外光谱结合化学计量学

摘要 本试验是为了控制龙胆的质量。通过近红外光谱。样品采集于整个青藏高原，其中生物活性成分为龙胆苦苷和龙根酸。使用浓度梯度法划分样品组。建模优化通过设计的正交测试 L9 (33) 进行处理。光谱预处理为乘法散射校正+二阶导数，采用偏最小二乘法建立模型。该模型显示出高 R 和低均方根误差。此外，残余预测偏差为 3.14。目前的定量模型可以快速准确地确定G.straminea的质量。