In order to guide the power flow calculation staff to identify and correct the ill-conditioned power flow, this paper summarizes the identification and correction methods of ill-conditioned power flow in the large scale power grid. It analyses in detail six types of factors which affect the power flow convergence from the data rationality, the NR iteration initial value, and the strategies of DC iteration, node type conversion, and iteration step and unbalance allocation, summarizes four kinds of ill-conditioned power flow identification elements and indicators, and proposes two kinds of ill-conditioned power flow correction methods. The correction of ill-conditioned power flow must adopt the corresponding correction methods according to different ill-conditioned characteristics and the restriction factors for the convergence to determine the appropriate adjustment range and measures. Numerical results and analysis show that the voltage attenuation index using the NR iterative information can automatically identify the ill-conditioned, restrictive factors and appropriate adjustment range and measures to optimize implementation of ill-conditioned power flow correction. This work is supported by Science and Technology Project of State Grid Corporation of China “Study of Diagnosis Method for Large Power Network Simulation Convergence Performance”.