Newton–Raphson power flow with constant matrices: A comparison with decoupled power flow methods

Abstract

This paper presents an investigation of five methods, which use constant matrices for solving the power flow problem. There are two new methods and they are based on the Newton–Raphson method with constant matrices of conductance and susceptance. The two aforesaid proposed methods are based on a decoupling principle, and the voltage angles and voltage magnitudes are calculated in decoupled forms. The other methods used, are XB, BX and primal. The results are compared on the basis of the convergence characteristics, number of iterations, memory requirements and the CPU times. This paper gives details of the present method’s performance in a series of practical problems on normal r/x ratio systems and also on high r/x ratio systems. The number of iterations and convergence characteristics of the two proposed methods, present a better performance than the decoupled versions XB, BX and primal.