Abstract: For data with the characteristics of "small sample, poor information," the concept of grey panel data is proposed, and a clustering method with the grey panel data based on the possibility function is developed. First, the application range of the possibility function is expanded, and a possibility function based on the interval grey number is derived. Afterward, aiming at the problem that the existing model cannot be applied to the grey panel data directly, a method for determining cluster weights of the grey panel data is proposed. According to the principle of new information priority, the grey cluster coefficient matrix at different moments is integrated to obtain the comprehensive grey cluster coefficient matrix. Finally, the objects are divided into grey categories under the grey cluster coefficient maximization principle. The proposed method is applied to the air quality assessment of eight major cities in China. Compared with the traditional panel data clustering method, it is found that the proposed method can refine and stratify the quality of the clustering results, which can make the clustering results clearer and easier to understand.

Key words: Interval Grey Number, Grey Clustering, Possibility Function, Grey Cluster Coefficient