College students are an important talent resource of the country, and their physical quality is the foundation of all qualities, which not only directly affects their own healthy growth, but also affects their ability to serve the society in the future. Therefore, the cluster extraction algorithm is introduced, and five cluster features, such as body mass index, vital capacity, endurance quality, flexibility strength quality and speed dexterity quality, are used to carry out fuzzy cluster analysis on college students' physical quality. The extracted correlation dimension is taken as the mixed pool feature of big data clustering, and the big data is clustered according to the extracted chaotic correlation dimension feature. Examples show that the results have important reference value for helping the all-round development of students' quality, optimizing education and teaching plans and perfecting talent training programs in colleges and universities.