随着生活水平的提高, 人们越来越重视食品安全问题, 假酒是食品安全中无法绕开的一个话题, 不同品牌白酒的快速准确辨识, 对假酒识别和食品安全都具有较大的现实意义。基于此, 提出一种将XGBoost算法结合激光诱导荧光技术(Laser-induced fluorescence, LIF)的不同品牌白酒快速识别方法, 把激光诱导荧光技术用于采集白酒的荧光光谱, 然后将获取的白酒原始光谱数据用XGBoost算法识别。实验以40°和45°白酒为研究对象, 选取6种酒样, 每种酒样采集40组光谱, 随机选取30组用于XGBoost模型的训练, 剩余10组用于训练好的模型测试, 实验中, XGBoost算法的训练用时为0.172 s, 训练好的模型测试识别率为98.33%。实验结果表明, XGBoost算法结合激光诱导荧光技术可快速准确识别不同品牌白酒。

With the improvement of living standards, people pay more and more attention to food safety issues, and counterfeit wine is an inescapable topic in food safety. The rapid and accurate identification of different brands of wine has great reality for counterfeit wine identification and food safety. Based on this, a method for rapid identification of different brands of white wine by XGBoost algorithm combined with laser-induced fluorescence (LIF) was proposed. The laser-induced fluorescence technique was used to collect the fluorescence spectrum of liquor, and then the raw spectrum data of the obtained liquor was identified by XGBoost algorithm. The experiments took 40° and 45° liquor as the research object, selected 6 kinds of wine samples, collected 40 sets of spectra for each wine sample, randomly selected 30 sets for XGBoost model training, and the remaining 10 groups were used for training good model test. In the experiment, the training time of the XGBoost algorithm is 0.172 s, and the trained model test recognition rate is 98.33%. The experimental results show that the XGBoost algorithm combined with laser-induced fluorescence technology can quickly and accurately identify different brands of liquor.