Trade Study for Effect of Advanced Control on Energy Storage Sizing for Ship Power Systems
Abstract
In the proposed work, we will utilize Model Predictive Control (MPC) which is a form of recursive on-line optimization well suited to managing system constraints. This controller will be used to match power available in the network with demand loads. Operating over a given mission profile, the MPC will utilize both generation and storage to meet the set of loads. The knowledge of theupcoming mission loads allows the MPC to utilize preview optimization over a finite horizon and coordinate the use of power sources. The overall goal is to determine how much storage would be necessary to achieve a specified level of performance that could, for example, be characterized as maximizing power quality or minimizing voltage droop in the presence of disturbance loads. Akey is the prediction that is based on both a system model as well as knowledge of the upcoming loading cycle.We will start with a candidate ship power system topology provided by partners at the Naval Surface Warfare Center (NSWC). The assumption is that the number of elements do not change and the primary adjustment variables are (i) the size of the storage elements, (ii) loading cycles and (iii) the uncertainty in both the plant models and the loading cycles. We will construct a dynamic simulation model using available tools: e.g., SimPower, Opal RT, or PowerFlow toolsets. From this simulation model we will extract a control oriented model suitable for an MPC approach. Compared with the power conversion and transmission, the conversion and storage elements are relatively slow. Therefore, a time scale separation may need to be incorporated into the model for the MPC. Utilizing a set of mission profiles, possibly scaled off of a baseline mission profile, we will evaluate the performance of the controlled system with particular attention paid to minimum and maximum state of charge on the battery system. The total storage, possibly heterogeneously distributed, will then be adjusted until the maximum and minimum fall within acceptable bounds.Over the range of cycles and cycle uncertainty, a range of battery storage (MJ or kW-hr) can be provided for designer choices that reflect confidence in a given estimate.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 07, 2019
- Source ID
- N000141912064
Entities
People
- Andrew G. Alleyne
Organizations
- Office of Naval Research
- United States Navy
- University of Illinois Urbana–Champaign