Abstract: The monitoring data of power grid has characteristics of multiple sources,high dimensional,prior and hierarchical structure.Traditional power grid monitoring operation analysis technologies relying on artificial experience have been unable to meet the development requirements of centralized control and integration operation of bulk power system.It is necessary to achieve efficient collection and information mining in the power grid monitoring,to improve the ability of dispatchers to actively detect the real-time operation status of the power grid.By analyzing the limitations of the existing business system in terms of data specification and analysis methods,this paper illustrates the necessity of the construction of the monitoring system of smart grid operation data.At the same time,this paper puts forward to the technology line of data to model and model to application based on monitoring analyzing requirements.On the other hand,this paper designs the overall framework covering the bottom-up application of data access to business,and establishes a statistical analysis center,trend warning center,intelligent search,visualization exhibition center.It has made a breakthrough in standardizing data access,the whole process of data processing and the causal modeling technology,and a management model has been formed,which is made up of the formation of abnormal trend prior warning,rapid response and post analysis in closed loop.This paper introduces the effects of standardized data processing,business modeling and monitoring management on the basis of the construction and operation status of the demonstration project.