An efficient operation strategy for dispersed generation sources in bipolar asymmetric DC distribution networks: a sequential quadratic approximation

This paper proposes a sequential quadratic optimization of the optimal power flow (OPF) in bipolar direct current (DC) grids. This formulation is based on Taylor’s expansion applied to the non-convex constraints, thus transforming them into affine equations. This approach, suitable for both radial and meshed grids, considers that the neutral terminal is only grounded at the substation bus. Other groundings can be considered in the loads without a loss of generality. Two test feeders composed of 21 and 33 nodes are considered in order to validate the effectiveness of the proposed sequential quadratic convex approximation model. Since this approach is based on convex optimization, a fast convergence, the uniqueness of the solution, and the global optimum are ensured. Simulations were performed using Python with the CvxPy library, a modeling system specialized in convex programming, as well as the ECOS solver. The 21-bus grid was employed to validate the effectiveness of the proposed convex model regarding power losses minimization, and the 33-bus one was used to evaluate the effect of the efficient dispatch of renewable generators within day-ahead operation environments.