Feasible Point Pursuit Based Sequential Convex Approximation of Optimal Power Flow

Optimal Power Flow (OPF) is of paramount importance for the operation of the power network. However, the non-convexity of ACOPF poses convergence issues and limits its application in day to day operation. Therefore, a convex model is needed that gives guaranteed convergence and feasible solution. Among the existing convex OPF models, Second-Order Conic Programming (SOCP) is more efficient. However, the exact SOCP formulation of ACOPF has non-convex arctangent constraints. These constraints are generally either ignored or approximated to solve the relaxed and convex SOCP formulation of ACOPF. Consequently, retrieving a unique and feasible set of bus voltage phasors for ACOPF of meshed power networks is a challenge. Thus, this paper proposes a novel approach to recover a unique and feasible ACOPF solution from its convex formulation. The alternate difference of convex functions is used to reformulate the arctangent constraints. The proposed formulation is solved by a modified inner approximation technique such as Feasible Point Pursuit based Sequential Convex Approximation (FPP-SCA) to preserve the feasibility of the solution.