When the People's Bank of China formulates monetary policy, the issue of the structural characteristics of the persistence of domestic inflation is one of its important considerations. The traditional univariate unit root test method is often due to the inherent heterogeneity between panel data and The cross-section correlation problem results in unsatisfactory inspection results. In this paper, the panel data SURADF method is used. Under the premise of the correlation and data heterogeneity of different types of commodity inflation in China, the structural characteristics of the sustainability of the eight major categories of commodity inflation are divided into three regions: national, urban, and rural. Empirical research. The results show that there are significant differences in the persistence of inflation between different regions and between different categories of commodities. It can be seen from the comparison that the SURADF method is more accurate and more accurate than the traditional inspection method to show the persistence of inflation among different categories of commodities Sexual structural characteristics. It also proposes that the central bank should formulate more differentiated monetary control policies based on the different persistent structural characteristics of inflation in different regions and different types of commodities.