为保护牦牛肉在市场中的独特性和真实性,维护消费者的合法权益,本研究以近红外光谱技术(near infrared spectroscopy,NIR)对不同饲养模式的牦牛肉样品在全波近红外Whole-NIR(400～2 500 nm)光谱范围内结合簇类独立软模式法(Soft Independent Modeling of Class Analogies,SIMCA)建立模型来判别牦牛肉舍饲和放养来源的可行性.结果表明:在Whole-NIR(400～2 500 nm)光谱范围内原始光谱差异不明显,当因子数为5 时预测误差平方加和(PRESS)趋于平稳,最终无限趋近于0,NIR结合SIMCA模式识别方法得到Q-T2 分布图,类与类之间界限明显,能够将牦牛肉样品按照不同饲养模式分开聚类,模型性能优良,预测准确率高达 100%.综上所述,利用NIR结合SIMCA可实现对不同饲养模式牦牛肉真实性进行鉴别的目的.
Study on the Discrimination of Stall-Fed and Free-Range Yak Meat Using Near-Infrared Spectroscopy Combined with the SIMCA Algorithm
To protect the uniqueness and authenticity of yak meat in the market and safeguard the legitimate rights and interests of consumers,this study employs near-infrared spectroscopy(NIR)technology to analyze yak meat samples from different feeding regimes within the Whole-NIR(400～2 500 nm)spectral range,in conjunc-tion with Soft Independent Modeling of Class Analogies(SIMCA),to establish a model for discriminating between stall-fed and free-range yak meat sources.The results indicate that the original spectral differences within the Whole-NIR(400～2 500 nm)spectral range are not significant,but when the number of factors is set to 5,the predictive error sum of squares(PRESS)tends to stabilize and eventually approaches zero.The combination of Near-Infrared Spectroscopy(NIR)with SIMCA pattern recognition method yields a Q-T2 distribution plot,where clear boundaries between classes are evident.This enables the clustering of yak meat samples based on dif-ferent feeding modes,with excellent model performance and a high prediction accuracy of 100%.In summary,NIR combined with SIMCA can effectively discriminate the authenticity of yak meat from different feeding regimes.
Yak MeatNear infrared spectroscopySoft independent modeling of class analogiesHusbandry pattern
万方数据
维普
