TY  - EJOU
AU  - Li, Wenbiao 
AU  - Cao, Zhichong 
AU  - Li, Zhengyu 
AU  - Tao, Wenbiao 
AU  - Liu, Cheng 
AU  - Shi, Yuxin 
AU  - Tian, Rundong 

TI  - Energy-Based Approach for Short-Term Voltage Stability Analysis and Assessment
T2  - Energy Engineering

PY  - 2025
VL  - 122
IS  - 11
SN  - 1546-0118

AB  - With the increasing penetration of renewable energy in power systems, grid structures and operational paradigms are undergoing profound transformations. When subjected to disturbances, the interaction between power electronic devices and dynamic loads introduces strongly nonlinear dynamic characteristics in grid voltage responses, posing significant threats to system security and stability. To achieve reliable short-term voltage stability assessment under large-scale renewable integration, this paper innovatively proposes a response-driven online assessment method based on energy function theory. First, energy modeling of system components is performed based on energy function theory, followed by analysis of energy interaction mechanisms during voltage instability. To address the challenge of traditional energy functions in online applications, a convolutional neural network-long short-term memory (CNN-LSTM) hybrid artificial Intelligence approach is introduced. By quantifying the contribution of each energy component to voltage stability, key energy terms are identified. The measurable electrical quantities corresponding to these key energies serve as inputs, while the energy at the voltage unstable equilibrium point (UEP) obtained from offline simulations is used as both the energy threshold and the output of the artificial intelligence model, enabling the construction of an artificial intelligence model for energy threshold prediction. The measurable electrical quantities corresponding to these key energies serve as inputs, while the energy at the unstable equilibrium point (UEP) obtained from offline simulations acts as the output, enabling the construction of an artificial intelligence model for energy threshold prediction. Real-time response data are fed into the model to predict the system’s instantaneous energy threshold, which is then compared with the transient energy at fault clearance to evaluate stability. Validation on both a 3-machine, 10-bus system and the New England 10-machine, 39-bus system confirms the method’s adaptability and accuracy. The simulation results demonstrate that the proposed short-term voltage stability assessment model outperforms other methods in both accuracy and computational efficiency.
KW  - Short-term voltage stability; renewable energy; energy function; artificial intelligence

DO  - 10.32604/ee.2025.068683