Investigation on voltage stability evaluation indicators and algorithms for power systems based on neural network algorithms

In order to avoid the occurrence of unstable phenomena in the Power Grid (PG), it is necessary to have a thorough understanding of the current operating status of the PG. Research on voltage stability evaluation indicators and algorithms in the power system can assist power system operators in making effective decisions. In this paper, an improved continuous power flow scheme based on Lagrange quadratic interpolation was adopted. Secant method and Lagrange quadratic interpolation method were used for prediction comparison, and Newton method was used to correct the predicted results. This article utilized an online voltage safety assessment and warning model to ensure the safe operation of the PG and reduce the occurrence of faults. To ensure the integrity and accuracy of network dynamic information, a least squares fitting method based on orthogonal functions was adopted to improve the fitting accuracy without increasing communication volume. As the step size increases from 2 to 3, the number of power flow solutions decreased and the solution time increased. The effectiveness of the voltage stability evaluation algorithm in this paper was verified through simulation of two node systems: New England 39 and IEEE118.