Abstract
As a critical step to efficiently design control structures, system identification is concerned with building models of dynamical systems from observed input–output data. In this paper, a number of regression techniques are used for black-box marine system identification of a scale ship. Unlike other works that train the models using specific manoeuvres, in this work the data have been collected from several random manoeuvres and trajectories. Therefore, the aim is to develop general and robust mathematical models using real experimental data from random movements. The techniques used in this work are ridge, kernel ridge and symbolic regression, and the results show that machine learning techniques are robust approaches to model surface marine vehicles, even providing interpretable results in closed form equations using techniques such as symbolic regression.
Highlights
System identification, known in industrial design as surrogate modelling, is one of the most important phases in multiple engineering areas, where reliable mathematical models, and tools are needed for a wide range of applications [1,2]
This paper showed alternatives for black-box marine identification using different regression techniques
The standard RR has been treated as a particular case of polynomial kernel with p = 1
Summary
Known in industrial design as surrogate modelling, is one of the most important phases in multiple engineering areas, where reliable mathematical models, and tools are needed for a wide range of applications [1,2]. In [34] a surface marine vehicle is modelled using Kernel Ridge Regression with Confidence Machine (KRR-CM) They use a small number of basic and simple trajectories to train and test the models. Unlike other studies in the state-of-the-art that employ synthetic data without noise or small experimental datasets of basic movements, the work described in this paper uses experimental data from different random manoeuvres and trajectories. In these previous works, models have been trained using specific trajectories, such as evolution circles or Zig Zags, and have been tested on the same type of movement.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
