Effective Coordination of Highly Heterogeneous Subsystems in a PEPDS
Abstract
The Power Electronic Power Distribution Systems (PEPDS) will revolutionize the design, operation, and control of high-performance microgrids, such as the Naval Shipboard Power System (SPS). To unlock the full potentials of the PEPDS concept, significant developments of high-performance hardware and advanced algorithm/software are both needed. In addition to improving the performance of the intelligent Power Electronics Building Blocks (iPEBB) of the PEPDS, significant effort should also be placed on advanced control studies. A good control algorithm can improve systems performance, lower hardware requirement and cost, bring new functionalities, and give feedback to the hardware designs for further improvement.This proposal targets at effective coordination of multiple heterogeneous subsystems in a microgrid. It is known that a traditional Synchronous Generator (SG) has large capacity and slow response speed, which might be ten times different from that of the small and fast power electronics based Distributed Generators (DGs). Under existing control solutions, the fast and small DGs will soon hit their capacity bounds and lose effects. Afterwards, the slow and large DGs will dominate the dynamic process and result in a long settling time. The virtual synchronous generator (VSG) technique tries to slow down the response of fast DGs by emulating a slow SG. VSG can simplify control design but cause huge waste of the excellent capabilities of modern power electronic techniques. We plan to design advanced control algorithms for effective coordination of the heterogeneous subsystems in a PEPDS. First, dual-stage control will be introduced to achieve good steady state and dynamic control performance. Second, distributed consensus-based optimal control algorithms will be designed to further realize steady state optimality.Finally, deep reinforcement learning will be introduced to realize both dynamic and steady state optimality. The designed algorithms will significantlyimprove the energy efficiency, reliability, and survivability of the Naval SPS. To evaluate the performance of the designed algorithms, both switch-level model based simulations and hardware experimentations will be conducted. Up on the approval of a new DURIP award, simple Silicon Carbide (SiC) technology based iPEBB modules will be developed. The new iPEBB based PEPDS testbed will significantly improve the TRL of the research outcome.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 06, 2021
- Source ID
- N000142112165
Entities
People
- Wenxin Liu
Organizations
- Lehigh University
- Office of Naval Research
- United States Navy