Reproducibility Verification of Ship Propulsion System Model

The ship propulsion simulation models’ reproducibility in python

Master thesis (2019)

Authors

X. Wang Mechanical Engineering

Contributors

P. de Vos Ship Design, Production and Operations - Mechanical, Maritime and Materials Engineering (supervisor 1)
K. Visser Ship Design, Production and Operations - Mechanical, Maritime and Materials Engineering (supervisor 2)
P. de Vos Ship Design, Production and Operations - Mechanical, Maritime and Materials Engineering (supervisor 2)
M.B. Duinkerken Transport Engineering and Logistics - Mechanical, Maritime and Materials Engineering (supervisor 2)
Faculty
Mechanical Engineering
Copyright: © 2019 Xiaozhe Wang
More Info
expand_more

Published Date

14-05-2019

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Mechanical Engineering

Abstract

The ship propulsion system model is simulated by MATLAB/Simulink initially and is required to reproduce the simulation results in python. This master graduation project deals with the reproducibility of ship propulsion system model by different computer simulation tools. The MATLAB/Simulink model will be provided by Maritime Technology (MT) Department. The target is investigating and assessing the reproducibility of the ship propulsion model in python.
During the pre-study, the stiff problem occurs due to many differential equations in the ship propulsion system. So, the initial python model could use the wrong method to solve the stiff problem and the initial python model cannot use variable time step to simulate the model. Even if the model cannot reproduce some experiments. The project should find a new reproducibility assessment method in python.
The new reproducibility assessment method is first to analyze the model. This step can help us to know whether the simulation model is a stiff system or not. The report introduces a time-step test method for the stiff system with physical verification. At present, there is no exact definition of the stiff system. This method tries to find an effective way to test the stiff system.
Next, introduce two Ordinary Differential Equation (ODE) solver methods. One is the explicit method; the other is the implicit method. Previously, published research has found that the implicit method can solve the stiff problem effectively. Therefore, the ODE solver is the key to the propulsion system model.
Finally, the python modelling method needs to be improved, because the initial version model in python only uses fixed-time step. The new modelling method can apply a variable time step for the propulsion model. It will help to test the stiff system and save more computation resource.
Once the new python model is finished, by comparing these two methods in MATLAB/Simulink and python, the plot presents the simulation result which is accurate and close to the reference result. The verification use Mean Squared Error (MSE) to indicate the reproducibility result. Each sample point of the two simulation tools is compared. The error plot will present what the difference between the python model with MATLAB/Simulink model. If the MSE value is close to zero, the reproducibility result will be successful.
Finally, this graduation project gives the solution for the reproducibility assessment method in python. Study the stiff system solution for marine engineering, achieve the purpose of the reproducibility research and successfully reproduce the ship propulsion simulation model in python.