The evaluation of transient voltage stability of power system is one of the foundations to support the safe

stable and efficient operation of the system. In the case of a high proportion of renewable energy access

the randomness and complexity of the power grid have increased significantly

the existing model-driven and data-driven transient voltage discrimination methods are difficult to model and lack mechanism support

and all the shortcomings above become increasingly prominent. In view of the shortcomings of the above methods

this paper proposes a transient voltage stability evaluation method based on the maximum Lyapunov expendent index (MLE) of the network inactive quantity

which realizes the fusion of model-driven and data-driven advantages. Firstly

the network energy function based on response information is constructed by integrating the response trajectory of the system. Then

the correlation coefficient between voltage and network functional quantity and network non-functional quantity is defined

and then the network non-functional quantity is extracted as the voltage dominant term. Furthermore

according to the intrinsic relationship between the transient voltage stability and the amplitude change rate of the network inactive quantity and the difference between the network inactive quantity and the phase space reconstruction theory

the transient voltage stability problem is transformed into the MLE curve analysis problem of the network reactive energy difference of the key branch where the voltage disturbed node is located to realize the transient voltage stability discrimination of the system. Finally

the accuracy and effectiveness of the proposed method are verified by the simulation analysis of the modified New England 10-machine 39-node case and the electromechanical transient standard case CSEE-VS.