ABSTRACT
Objective::To study the quality of Curcumae Radix from different areas, in order to screen out excellent provenances to improve the overall quality and stability of Curcumae Radix. Method::The volatile oil, alcohol extract, germacrone, germacr-1(10)-ene-5, 8-dione and curcumin were selected as comprehensive evaluation indicators to determine the content of five main components in 50 samples of Curcumae Radix from different areas. Cluster analysis was made for Curcumae Radix from different areas of Guangxi, and the fusion models of GRA(grey correlation degree) and TOPSIS (technique for order preference by similarity to ideal solution) of Curcumae Radix were established. Result::The results of cluster analysis indicated that 50 samples from different areas could be divided into four categories. The fusion models of GRA and TOPSIS were based on relative closeness γi as the index to comprehensively evaluated the quality of Curcumae Radix. Samples No. YL-2, YL-5, YL-1, YL-4 were screened out as excellent samples. The average relative closeness γi was ranked as follows: Yulin>Pingnan>Hengxian>Xingye>Zhaoqing>Qingtang>Zhongshan>Qinzhou>Guigang>Lingshan. Conclusion::The multiple attribute decision making analysis gray correlation degree and TOPSIS fusion models are used to optimize the quality evaluation of Curcumae Radix. The method is simple, objective and comprehensive, which can be promoted and provide a reference for the screening of high-quality provenance.