模式识别技术用于果酸混合物分析

Pattern Recognition Techniques Applied to Analysis of Fruit Acid Mixture

研究了偏最小二乘法（ＰＬＳ）和人工神经网络法（ＡＮＮ）用于水果的紫外可见多组分光度分析。当混合物中诸组分存在相互作用，光谱加和性受到扰动时，ＰＬＳ法和ＡＮＮ法用于混和物的分析仍能得到较满意的结果。

Direct quantitative determination of multicomponents of fruit have been developed .The substances to be analysis were malic acid,tartaric acid,citric acid and oxalic acid.There is interaction between these substances and the spectral idditivity is violated.For resolving these spectra and quantifying these substances,two pattern recognition techniques ,the partial least square(PLS) and artificial neural network(ANN) were applied and satisfactory results were obtained.