A hybrid decoupled approach to optimal power flow

This paper presents a hybrid decoupled optimal power flow approach aimed at implementing an optimal power flow program for online applications. By taking advantage of the partial duality principle of mathematical programming theory and the inherent weak coupling characteristics between real and reactive power flows in a power system, the solution to the original optimization problem can be approximated iteratively by solving a real and a reactive subproblem alternately. Two special deliberately customized techniques, one (based on linear programming) and the other (based on quadratic programming), are adopted to solve the two subproblems respectively. This hybrid feature makes the approach very flexible while preserving the efficiency of decoupling. The approach has been successfully implemented for online usage in the Control Center of Northeast China Electric Power Administration