Abstract
Generally, multi-terminal dc grids provide full control over the dc powers that converter terminals exchange with their respective host ac grids. Nevertheless, the same level of controllability does not exist over the power flow in the individual dc lines of a highly meshed HVDC network, and this increases the risk of overloading the dc cables that present lower resistances. To address the highlighted shortcoming, this paper presents a generic dc grid controller that uses nonlinear constrained optimization to optimize the performance of multi-terminal HVDC networks for any desirable operational objective, within system physical constraints such as dc cable and converter thermal limits, and minimum and maximum dc voltage limits. The presented dc grid controller performs online optimization at regular intervals to dynamically estimate and update the set-points of the converter terminals as system operating conditions vary. The technical viability of the proposed controller is assessed using a generic seven-terminal HVDC network that uses voltage sourced modular multilevel converters. Simulations from the scenarios that prioritize dc grid power loss or operational cost minimization show that the presented dc grid controller exhibits good performance during normal operation as converter terminals set-points vary, and during dc grid reconfiguration following simulated successive outages of the dc cables. It has been shown that the performance of the proposed DC grid control is not affected by the parameter variation such as changes of resistance with temperature.