提出了一种对多组分混合物的紫外连续光谱进行数据处理实现定量分析的方法.该方法将待测浓度和吸光系数同时作为决策变量,把分析问题表示为一个有约束优化问题,建立多组分定量分析数学模型,采用广义简约梯度算法求解。通过对典型的2-5组分食用色素混合物的紫外连续光谱数据进行分析,分析结果是平均和最大相对误差分别为2.075%、7.800%,说明本文方法具有较高的准确度,是一种方便、准确的多组分定量分析方法。 A method for quantitative analysis of ultraviolet continuous spectrum of multi-component mixtures was proposed, in which the concentration to be tested and the extinction coefficient were taken as decision variables at the same time, the analysis problem was expressed as a constrained optimization problem, and multiple groups Quantitative analysis of mathematical models, using generalized simple gradient algorithm. By analyzing the UV continuous spectral data of typical 2-5 component food color mixture, the average and relative maximum relative errors are 2.075% and 7.800% respectively, which shows that this method has high accuracy and is convenient , Accurate multi-component quantitative analysis.