Natural frequency is one of the main dynamic characteristics of damage diagnosis methods, which is not only related to the stiffness change of the structure itself, but also easily affected by environmental factors such as temperature. A natural frequency clustering analysis method is proposed, which is used for damage diagnosis of space grid structure under environmental changes. First, the influence of the natural frequency of the structure on the environmental temperature is derived. based on this analysis, a grid is used as the research object to simulate the natural frequency data under the condition of temperature changes and damage; then the principal component analysis (PCA) of the natural frequency is developed, and the structural damage diagnosis was realized by using fuzzy c-mean(FCM) clustering based on principal component reconstruction residuals. The research results show that the ambient temperature affects the natural frequency by causing thermal deformation,changing the material properties and generating temperature internal forces; the daily fluctuation of the naturalfrequency with temperature will directly affect the accuracy of damage diagnosis, and the PCA-FCM clustering can be used in the damage diagnosis process which can effectively eliminate the interference of environmental temperature factors, judge whether the damage occurs when the health baseline is unknown, and accurately judge the degree of damage.