Shipboard electrical system modeling for early-stage design space exploration

Abstract

In early-stage design exploration, it has been found that electrical dynamics do not significantly affect the dependability of an integrated engineering plant. Therefore, it has been found useful to neglect these electrical dynamics and focus on mechanical, thermal, and fluidic dynamics in assessing system performance. Previous methods of accomplishing this goal involve the use of linear programming to describe the behavior of the electrical system. Herein, two significant shortcomings of the existing linear programming methods are identified, and a method of representing the electrical system that addresses these shortcomings is proposed. The proposed method is demonstrated in several system studies.

Upper Bound Performance of Shipboard Power and Energy Systems for Early-Stage Design

“…Allowing such a large prediction horizon is an optimistic assumption, but it does provide a rigorous upper bound on system performance and avoids the difficulties of selecting among thousands of possible controller designs. The results can be compared against the performance of a controller with no future knowledge [19], [20], [23].…”

“…The constraints g and h include the network flow constraints as described later in Section II-C. The constraints (4) are of the form (19) to impose limits on the voltage magnitude and angle at each bus and the minimum and maximum real and reactive powers of each generator and energy storage unit.…”

Expanding the Design Space Explored by S3D

“…It is possible to very rapidly determine the amount of power flowing through each component in a system for a given operational alignment by linearizing the components and solving the resultant system of linear equations. Examples of this process can be found in [11] and [12]. However, as noted above, we must determine the flow under any possible alignment.…”

The role of future information in control system design for shipboard power systems

“…With some simplifying assumptions, the problem can become a linear programming problem. More details of the actual numerical solution method used here are available in Cramer et al (2013Cramer et al ( , 2015.…”

“…The multi-period controller is very difficult to fully implement in practice as it assumes exact future knowledge on a long horizion, but it provides an upper bound on the system performance given a specific system architecture. These two controller types have been proposed separately before, where the instantaneous version was discussed in Chan et al (2009);Cramer et al (2013Cramer et al ( , 2015 and the multi-period version was shown in Oh et al (2017). The goal here is to compare the two and rigorously quantify the value of the forecasts for different architectures and operating conditions.…”