Project Description
The objectives of this research are to develop a Control Hardware-in-the-Loop (CHiL) emulation of a Robust Combat Power electrical distribution, power conversion and energy storage system in order to perform metrics during a range of survivability events as a function of energy storage and energy storage management controls. This work is part of a multi-university effort to develop CHiL based emulations of future naval combat systems with mixed AC and DC distribution. UWM's role in this effort is to simulate the details of protection hardware and protection schemes and to determine the levels of detail required to emulate system invulnerability and recoverability to a set of survivability scenarios. Survivability scenarios include loss of generation and line-to-line short circuit faults at various locations within a multi-zonal distribution system. UWM's task is to develop the multi-zonal simulations with protection mechanisms, perform fault characterization and provide metrics that can be used to assess measures of effectiveness. This work is core to UWM's research contribution overall to the navy's ship electrification programs. It provides a showpiece for distributed control and protection capabilities that enhance the survivability of such systems. The aim is to encourage adoption of these distributed control and protection capabilities.
Tasks and Responsibilites
This student will lead efforts to develop the multi-zonal shipboard simulation with advisement from two PhD students. There are a number of models for power and energy conversion that have been developed by contributors to the navy's Robust Combat Power Challenge (RCPC) program. The bulk of these models have been developed at John Hopkins University. The student will assess the effectiveness of these models for fault protection and loss of generation simulations and make necessary improvements to the models. It is expected that the models provided, which are average models, will be insufficient. Full switching models must be developed to determine the response to faults and, where it is possible, these switching models will inform adaptations to the average models. In some cases, full switching models are necessary but it is recognized that full switching models throughout is impractical.