Power grid transient stability prediction method based on power angle maximum Lyapunov index trajectory reasoning

The invention discloses a power grid transient stability prediction method based on power angle maximum Lyapunov index trajectory reasoning. A phasor measurement unit (PMU) in a power system is used for monitoring transient response of the system after a fault occurs; according to the power angle of the generator, calculating a power angle MLE track of a transient state last phase; constructing and training an MLE trajectory reasoning model and reasoning a power angle MLE trajectory by adopting a GRU algorithm in combination with transient initial stage data and a transient final stage power angle MLE trajectory in the transient response information; in combination with the inferred power angle MLE trajectory and the corresponding transient stability category, an XGBoost algorithm is adopted to construct and train a transient stability discrimination model; when the method is applied to an actual system, after a fault occurs, transient initial data in transient response information is input into the MLE traje