Soft independent modeling class analogy (SIMCA) was applied to infrared spectra, obtained by attenuated total reflection – Fourier transform infrared (ATR - FTIR) spectroscopy, to develop discrimination models for the floral and geographical origin of honeys. Mono-floral honey samples, originated from France and Romania, were analyzed for this purpose. The meaningful spectral areas which were selected for the development of discriminating models were: 500 - 1610 and 1700 - 3000 cm⁻¹. A second derivative and an “autoscale” processing was performed to experimental spectra before the statistical classifications. The honeys varieties were accurately classified, independently of the geographical origin of honeys, suggesting the high potential of this technique for botanical recognition. The obtained interclass distances among honey varieties were significant and the main identified discrimination markers, from FTIR spectra, are especially related to the floral origin through the pollen gathered by bees. A global geographical classification of honeys provided a correct classification with a weak interclass distance showing the model limitations for the authenticity control purposes. Nonetheless, geographical classifications inside a single class of honey variety lead to a correct classification with a significant interclass distance.

将软独立建模类比法（SIMCA）应用于通过衰减全反射–傅里叶变换红外（ATR-FTIR）光谱法获得的红外光谱，以开发针对蜂蜜的花香和地理起源的判别模型。为此，对源自法国和罗马尼亚的单花蜂蜜样品进行了分析。为开发区分模型而选择的有意义的光谱区域为：500-1610和1700-3000 cm ^-1。在进行统计分类之前，对实验光谱进行了二阶导数和“自动缩放”处理。准确地对蜂蜜品种进行了分类，而与蜂蜜的地理来源无关，这表明该技术在植物学识别方面具有很高的潜力。蜂蜜品种之间的类间距离非常重要，从FTIR光谱中识别出的主要判别标记尤其与蜜蜂采集的花粉引起的花起源有关。蜂蜜的全球地理分类提供了正确的分类，其类间距离较弱，显示了出于真实性控制目的的模型限制。尽管如此，