The Multi-Zonal Medium Voltage DC (MVDC) Shipboard Power System (SPS) architecture, proposed by the U.S. Navy for their future combatant system, consists of several voltage source converters (VSCs). The proposed architecture is tightly-coupled, power-limited and its performance needs to be evaluated for security, reliability, and survivability. Following system damage or a fault, the current flow pattern in the DC network may change, which may result in the failure of VSCs due to overvoltage developed across them in certain operating conditions. For a given MVDC system, DC voltage reference setting for one of the VSCs operating in the voltage regulator mode, and the optimal power reference settings of the remaining VSCs in the power dispatcher mode have to be pre-determined. These settings and control modes of VSCs are needed to maintain the DC voltage within desired margins (usually 5% around the nominal DC voltage), both in “pre-fault” and “post-fault outage” conditions. The problem has been formulated as an optimization problem with three different objective functions. Computational intelligence techniques have been applied for solving the optimization problem. These include the genetic algorithm (GA) and biogeography based optimization (BBO) methods. The results have been compared with a conventional Lagrange multiplier based method.
展开▼
