The real-time assessment and prediction of the carbon intensity level of the distribution system is an important part of the low-carbon optimal operation of the energy flow-carbon flow on the distribution side. In order to quantify the carbon emission sources and carbon control margins of the distribution system

a generalized carbon intensity awareness method for the distribution network based on carbon emission flow is proposed. Firstly

on the basis of the carbon emission flow theory

the generalized carbon intensity of the distribution network considering the influence of active network loss is calculated by using the operating situation elements and carbon intensity elements of the distribution network

which reflects the spatial difference of carbon intensity of nodes in the radial distribution network. Secondly

in view of the limited sample size of carbon intensity data and the variability and complexity of source-load information

a multi-variate Padé approximation dendrite net (multi Padé-DDNet) prediction model is established to achieve high-precision generalized carbon intensity prediction results by strengthening the nonlinear mapping relationship between carbon intensity-related elements and model output. Thirdly

the nodal carbon intensity margin index

branch carbon intensity margin index

and network-level carbon emission margin index are established

progressing hierarchically from the key nodes to subdivisions and then to the distribution network. The high carbon intensity thresholds of different regions and different particle sizes in the distribution network are combined to perceive the system security situation operational level and dynamic high carbon risk. Finally

the feasibility and effectiveness of the proposed method are verified by taking an actual distribution network in China as an example.