红参Ginseng Rubrum具有抗癌、抗衰老、降血糖、降血压、提高心血管功能等药用价值，然而，市场上红参质量参差不齐，给消费者的使用以及工业加工生产带来困扰。因此，对红参原料药材快速、准确的定性和定量分析对于红参质量控制显得尤为重要。近红外光谱能够反映被测对象组织结构和成分信息，随着科学技术的发展和分析技术的进步，近红外光谱（NIR）技术以其快速、便捷的优势被广泛应用于红参鉴别、红参化学成分含量检测以及红参工业加工过程控制的研究中。然而，红参化学成分的复杂性和药理作用的不确定性，以及NIR分析建模的样本数量和多样性及数据采集过程中设备、环境、样本非目标成分的影响等降低了模型鲁棒性，限制了该方法在实际中的推广应用，因此如何在数据采集部分对系统进行优化以及在信息处理部分降低非目标成分的影响来提高信噪比，增强模型预测能力还有待研究。

Red ginseng exhibits a variety of pharmacological activities, including anticancer, anti-aging, anti-diabetics, anti-hypertensive, which can enhance cardiovascular function effects.However, the quality of red ginseng is uneven in the market, and causing trouble to consumer and industrial production.Therefore, qualitative and quantitative analysis is vital to guaranteeing the quality of red ginseng.Near infrared spectroscopy (NIR) has been used for the identification, chemical content, and procedure control of red ginseng.However, beyond the existence of complex chemistry content and uncertainty pharmacological action of red ginseng, NIR model robustness is influenced by equipment, environment and nontarget component of sample itself, and primarily to limit its popularization and application of actual industry.How to optimize data acquisition and procession to improve signal to noise ratio and enhance prediction ability of model still needs to be studied.

国家自然科学基金资助项目（81573826，81202643，81774148）；天津市科技计划项目（15PTCYSY00030）；国家中药标准化项目（ZYBZH-C-JS-32）