Multiobjective Optimization of Multiconductor DC Power Cables

In this paper, an equivalent circuit approach is proposed to model the electric and thermal responses in multi-conductor DC cable systems intended for electric ships. The proposed model is sufficient to address the complexity of the nonlinear physics, including space charge accumulation within the dielectric, yet flexible enough to enable rigorous population-based design. Indeed, a genetic algorithm has been implemented to establish the Pareto-optimal fronts between mass, size and loss for cables of variable geometry and conductor material properties. The fronts of the multi-conductor cable are then compared to those of single conductor cables.