▎ 摘　　要

NOVELTY - The method involves dividing complex track into multiple weeks from outer to inner, and defining an identification characteristic set according to the number of weeks divided on the complex track for subsequent big data analysis algorithm. The identification characteristic set is extracted according to a certain number of known characteristic peak samples, and utilizing the extracted characteristic as input data to train an intelligent algorithm for establishing a graphene Raman spectrum automatic identification model based on big data. A graphene Raman spectrum to be identified is automatically identified according to the graphene Raman spectroscopy automatic identification model. Defect judgment and layer number judgment processes are performed according to an identification result. USE - Big data analysis based graphene fingerprint peaks analyzing method. ADVANTAGE - The method enables performing graphene single layer/multilayer process according to the Raman characteristic spectrum of graphene, or automatically identifying whether defect, and improving identification accuracy and efficiency. DESCRIPTION OF DRAWING(S) - The drawing shows a graph view of two peak neighborhoods.