Strictly Convex Loss Functions For Port-Hamiltonian Based Optimization Algorithm For Mtdc Networks

In this work we propose a primal-dual method that can be cast in a port-Hamiltonian framework for minimizing the power losses in a multi-terminal DC network. The main contribution consists of proposing an alternative power loss function by means of a change of variables that translates the convex objective function into a strictly convex objective function. The results hold under some restrictive assumptions, but necessary to make steps towards a complete algorithm in future research. The obtained results are validated via numerical simulations.