Modeling and simulation of an all electric ship in random seas

Abstract

This Masters thesis, conducted in support of the All Electric Ship (AES) early design effort, presents two computational programs for analysis and simulation: a full-scale, end-to-end AES simulator and an analytical performance and stability assessment tool for the ship's propulsion drive; the integrated power system (IPS). The AES simulator incorporates high order techniques for the hull modeling with low order, low effort models for the propellers, IPS, and prime movers, culminating in a fully-coupled, end-to-end, simulation environment, which is still practical for high effort studies like uncertainty quantification or optimization. The most appealing characteristic of this program is the time domain hull model with combines nonlinear maneuvering equations, seakeeping equations, and second order wave force equations. This allows for the prediction of propeller elevation and inflow velocity in random seas, and effectively the high fidelity modeling of propeller load schedules. This capability is vital for AES design where propeller load fluctuations can lead to large electrical power transients onboard. To demonstrate the capability of the AES simulator, ship trails are run in calm and random seas. IPS state evolutions are given to show the propagation of load disturbances. Monte Carlo methods are applied to assess transients in the inherently random sea environment. The IPS assessment tool attempts analytical quantification of the performance and stability of the Purdue MVDC Testbed, a scaled IPS composed of analagous elements: electric machinery, power converters, MVDC distribution, and bus voltage/induction motor torque control schemes. The thesis details the applicable nonlinear equations and the tools for identifying system equilibrium points. Then, small displacement theory is used to attain linear state space matrices valid near the operating points, from which traditional stability and performance techniques can be applied. Methods for closed loop analysis are suggested including ways to assess the hysteretic control elements used for induction motor torque control. Results from experiments with the high fidelity, high effort, Purude MVDC Testbed model are used for validation.