[Purpose/significance] Network public opinion leaders play an important role in spreading,guiding and promoting the development of emergencies. It is of great significance to effectively identify opinion leaders for monitoring and early warning of network public opinion. [Method/process] Combining with the propagation characteristics of network public opinion in the microblog platform,from the dimensions of impact,activity and identity,the paper builds a micro-blog opinion leaders index system,which determines the index weights by an improved fuzzy AHP method,and establishes an opinion leader identification model and influence ranking method with grey relational analysis method. [Result/conclusion] The paper provides an opinion leader identification model,which can scientifically and effectively identify and analyze the opinion leaders of network emergencies. [Limitations] There is a certain degree of scalability in the quantitative dimension of indexes and implicit index data.