TOPSIS与灰色关联分析在不同产地炒茺蔚子质量评价中的应用

TOPSIS and Grey Correlation Analysis on Quality of Stir-baked Leonuri Fructus from Different Producing Areas

目的:应用逼近理想排序法(TOPSIS)与灰色关联法评价不同批次炒茺蔚子的质量并进行比较。方法:测定不同批次炒茺蔚子中总灰分、酸不溶性灰分、浸出物、盐酸水苏碱的百分含量,通过层次分析法进行指标权重的确定,得到4个指标的权重为0.077,0.077,0.385,0.461;采用MATLAB编程构建不同产地炒茺蔚子的灰色关联度质量评价模型并与TOPSIS分析结果比较。结果:对不同批次炒茺蔚子中的各成分进行灰色关联分析,得出"广东20121117"的药材与最优参考序列的相对关联度最高(0.561 685 996)质量最优,"河南20130401"的药材与最优参考序列的相对关联度最低(0.402 108 175)质量最差,TOPSIS与灰色关联法的分析结果基本一致,说明结果的可重复性较好,较为可靠。结论:灰色模式识别模型,适宜对多组分多指标的中药材样品质量进行综合评价,具有推广应用前景。

Objective： To evaluate the quality of different batches of stir-baked Leonuri Fructus and make a comparison by using TOPSIS and grey correlation analysis. Method： The percentage composition of total ash,acid-insoluble ash,extractum and stachydrine hydrochloride in different batches of stir-baked Leonuri Fructus were determined. The analytic hierarchy process was used to determined the four index weights 0. 077,0. 077,0. 385,0. 461. MATLAB programming was adopted to build the grey correlation quality evaluation model for stirbaked Leonuri Fructus from different producing areas and compare the TOPSIS results. Result： According to the grey correlation analysis on components in the different batches of stir-baked Leonuri Fructus,Leonuri Fructus with the batch of ＂Guangdong 20121117 ＂ was the most highly correlated with the optimal reference sequence（ 0. 561 685 996） and so the best in quality; Leonuri Fructus with the batch of＂Henan 20130401＂was the most lowly correlated with the optimal reference sequence（ 0. 402 108 175） and so the worst in quality. The results are basically identical with that of TOPSIS,indicating a good repeatability and stability. Conclusion： The grey recognition model is applicable for the comprehensive quality evaluation of multiple-component and multi-target samples of traditional Chinese medicines,with a good promotion and application prospect.