Adaptive Learning of Inland Ship Power Propulsion under Environmental Disturbances

Nicolas Dann; Pablo Segovia; Vasso Reppa

doi:10.1016/j.ifacol.2022.10.400

Adaptive Learning of Inland Ship Power Propulsion under Environmental Disturbances

Nicolas Dann, Pablo Segovia, Vasso Reppa

Transport Engineering and Logistics

Research output: Contribution to journal › Conference article › Scientific › peer-review

39 Downloads (Pure)

Abstract

This paper presents an adaptive approximation-based scheme for learning a partially known ship power propulsion plant under various environmental conditions. Considering the effect of water depth on the engine power, a dynamic model is defined comprised of the engine dynamics and the 1-DoF ship manoeuvring dynamics. The modelling challenge is the determination of ship resistance. To meet this challenge analytical modelling of ship resistance is combined with an error-filtering online learning (EFOL) scheme for computing an approximation of the unmodeled part of ship resistance related to wind and air. After simulations under multiple weather conditions, the trained model was demonstrated to efficiently estimate the unmodelled part of the ship resistance for an inland vessel.

Original language	English
Pages (from-to)	1-6
Journal	IFAC-PapersOnline
Volume	55
Issue number	31
DOIs	https://doi.org/10.1016/j.ifacol.2022.10.400
Publication status	Published - 2022
Event	14th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles, CAMS 2022 - Kongens Lyngby, Germany Duration: 14 Sept 2022 → 16 Sept 2022

Keywords

On-line learning scheme
shallow water
ship resistance
speed-power prediction
surface vehicles

Access to Document

10.1016/j.ifacol.2022.10.400

1-s2.0-S2405896322024466-mainFinal published version, 903 KBLicence: CC BY-NC-ND

Cite this

@article{74b6bafc49bb4b35b98ddbb14ad834da,

title = "Adaptive Learning of Inland Ship Power Propulsion under Environmental Disturbances",

abstract = "This paper presents an adaptive approximation-based scheme for learning a partially known ship power propulsion plant under various environmental conditions. Considering the effect of water depth on the engine power, a dynamic model is defined comprised of the engine dynamics and the 1-DoF ship manoeuvring dynamics. The modelling challenge is the determination of ship resistance. To meet this challenge analytical modelling of ship resistance is combined with an error-filtering online learning (EFOL) scheme for computing an approximation of the unmodeled part of ship resistance related to wind and air. After simulations under multiple weather conditions, the trained model was demonstrated to efficiently estimate the unmodelled part of the ship resistance for an inland vessel.",

keywords = "On-line learning scheme, shallow water, ship resistance, speed-power prediction, surface vehicles",

author = "Nicolas Dann and Pablo Segovia and Vasso Reppa",

year = "2022",

doi = "10.1016/j.ifacol.2022.10.400",

language = "English",

volume = "55",

pages = "1--6",

journal = "IFAC-PapersOnline",

issn = "1474-6670",

publisher = "Elsevier",

number = "31",

note = "14th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles, CAMS 2022 ; Conference date: 14-09-2022 Through 16-09-2022",

}

TY - JOUR

T1 - Adaptive Learning of Inland Ship Power Propulsion under Environmental Disturbances

AU - Dann, Nicolas

AU - Segovia, Pablo

AU - Reppa, Vasso

PY - 2022

Y1 - 2022

N2 - This paper presents an adaptive approximation-based scheme for learning a partially known ship power propulsion plant under various environmental conditions. Considering the effect of water depth on the engine power, a dynamic model is defined comprised of the engine dynamics and the 1-DoF ship manoeuvring dynamics. The modelling challenge is the determination of ship resistance. To meet this challenge analytical modelling of ship resistance is combined with an error-filtering online learning (EFOL) scheme for computing an approximation of the unmodeled part of ship resistance related to wind and air. After simulations under multiple weather conditions, the trained model was demonstrated to efficiently estimate the unmodelled part of the ship resistance for an inland vessel.

AB - This paper presents an adaptive approximation-based scheme for learning a partially known ship power propulsion plant under various environmental conditions. Considering the effect of water depth on the engine power, a dynamic model is defined comprised of the engine dynamics and the 1-DoF ship manoeuvring dynamics. The modelling challenge is the determination of ship resistance. To meet this challenge analytical modelling of ship resistance is combined with an error-filtering online learning (EFOL) scheme for computing an approximation of the unmodeled part of ship resistance related to wind and air. After simulations under multiple weather conditions, the trained model was demonstrated to efficiently estimate the unmodelled part of the ship resistance for an inland vessel.

KW - On-line learning scheme

KW - shallow water

KW - ship resistance

KW - speed-power prediction

KW - surface vehicles

UR - http://www.scopus.com/inward/record.url?scp=85145657510&partnerID=8YFLogxK

U2 - 10.1016/j.ifacol.2022.10.400

DO - 10.1016/j.ifacol.2022.10.400

M3 - Conference article

AN - SCOPUS:85145657510

SN - 1474-6670

VL - 55

SP - 1

EP - 6

JO - IFAC-PapersOnline

JF - IFAC-PapersOnline

IS - 31

T2 - 14th IFAC Conference on Control Applications in Marine Systems, Robotics, and Vehicles, CAMS 2022

Y2 - 14 September 2022 through 16 September 2022

ER -

Adaptive Learning of Inland Ship Power Propulsion under Environmental Disturbances

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this