New methods for computing a saddle-node bifurcation point for voltage stability analysis

This paper proposes methods for calculating saddle-node bifurcation points of power system power flow equations. The first method calculates a saddle-node bifurcation point along a given ray in the parameter space of power flow equations. By exploiting the special structure of power flow equations, the method calculates a saddle-node bifurcation point along a given ray as a solution to a constrained optimization problem. The constrained optimization can be solved efficiently with standard optimization methods. The second method calculates a locally closest saddle-node bifurcation point with respect to the operating point. This method uses an iterative process of computing a saddle-node bifurcation point along a ray, and then updating the direction of the ray for calculating a closer saddle-node bifurcation point. The method is a quasi-Newton method that updates the direction of a ray based on the first-order derivatives and the approximations to the second-order derivatives of the distance between saddle-node bifurcation points and the operating point. The paper compares the proposed methods with other methods on two test examples