The solution of optimal power flow based on a semismooth subdifferential method

A semismooth subdifferential method (SSM) for solving optimal power flow (OPF) is presented. The Karush-Kuhn-Tucker (KKT) conditions of OPF system, by introducing nonlinear complementarity problem (NCP) function, are transformed equivalently to a set of semismooth nonlinear algebraic equations. Then an inexact Levenberg-Marquardt (L-M) algorithm based on the subdifferential is improved and employed to solve the equations. The positive definitiveness of the iterative coefficient matrix is enhanced by using L-M parameter, while a reformed nonmonotone line search is used to enforce global convergence. Finally, the effectiveness of the SSM is shown by analyzing the computational performance and convergence characteristic of the proposed algorithm on test systems.