With the increase of the proportion of new energy capacity

the uncertainty of the frequency dynamic process of the power system after large disturbance is obviously enhanced

the deterministic analysis method has been unable to accurately and comprehensively characterize the frequency safety state and evolution situation of the system

and the time-domain simulation method also can not meet the needs of system analysis because of its low analysis efficiency. For this reason

the solution idea of extreme point analysis is selected

taking the frequency nadir after disturbance as the analysis object

an analytical method for probability estimation is proposed: Firstly

the statistical eigenvalues describing the uncertainty of the influencing factors of the frequency nadir are obtained by data-driven and model analysis

which provide input for probability estimation; Then

the probability density function (PDF) transformation of multivariate function is carried out based on the deterministic frequency nadir analytical formula: The PDF of the time of frequency nadir is obtained by using the central limit theorem

then divided into independent and explicit coupling

using the way of marginalizing the joint PDF of equal-dimensional variable groups and the way of variable processing the analytical formula after dimensional compression

to obtain the frequency nadir deviation; And finally

the PDF of the frequency nadir value is obtained by convolution of the PDFs of the frequency nadir deviation and the initial frequency. The numerical results show that it is necessary to use probabilistic analysis for the frequency dynamics of the disturbed system under the present operating conditions

and the proposed method can significantly improve the analysis efficiency while ensuring the estimation accuracy compared with the Monte Carlo method

among them

the method without considering the coupling relationship between variables has higher analysis efficiency and is suitable for the case of independent variables

while the method considering coupling relationship has higher analysis accuracy in the case of explicit coupling between variables.