Non-parametric dynamic system identification of ships using multi-output Gaussian Processes

Ramirez, Wilmer Ariza; Leong, ZQ; Nguyen, Hung D.; Jayasinghe, Shantha Gamini

doi:10.1016/j.oceaneng.2018.07.056

Cited by 42 publications

(13 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The trajectory-tracking experience chosen in this work consists in evaluating the performance of a LQR based controller for tracking position zig-zag and circular trajectories in the horizontal plane, as they are usually employed by researchers in this kind experiments with ship model [12], [13], [14], [15], for the horizontal plane. To these ends, a linear model for the system must be computed.…”

Section: Methodsmentioning

confidence: 99%

Virtual environment for the design of position trajectory tracking controllers of remotely operated vehicles

Muñoz-Aldana

Gaviria-López

2019

ITECKNE

View full text Add to dashboard Cite

This article presents a virtual environment based on co-simulation between MatLab and MSC Adams, allowing simulation, analysis, development and validation of control strategies for tracking of position trajectories of a Remotely Operated Vehicle (ROV). The simulation results in the horizontal plane show that it is possible, in an uncomplicated way, to construct a virtual environment, which allows observing realistic movements when the forces exerted on an ROV are provided. Taking advantage of the properties of co-simulation, the experiences in this work show that this simulation strategy is very suitable for analysis purposes and control design, allowing researchers and professionals the wide use of control tools available in MATLAB for this end. In this work, a robust linear quadratic regulator (LQR) with integral action has been used to evaluate the performance of the proposed virtual environment for tracking of position trajectories. To validation purposes, widely used trajectories in naval study designs were employed such as the Zig -Zag shaped and the Circular shaped trajectories. Simulation results show that the integration of both, MatLab and MSC Adams, effectively addressees the problem of evaluation of performance of control strategies in the virtual environment. The presented approach allows gaining experience about the challenges of this kind of control problems, before dealing with the complex aspects of tuning in real experimental environments, avoiding losses and cost overruns for underwater robotics projects.

show abstract

Section: Methodsmentioning

confidence: 99%

Virtual environment for the design of position trajectory tracking controllers of remotely operated vehicles

Muñoz-Aldana

Gaviria-López

2019

ITECKNE

View full text Add to dashboard Cite

show abstract

“…The kernel methods overcome these problems based on statistical learning theory [35]. The kernel methods, such as SVM [28], the Gaussian process (GP) [36], locally weighted learning (LWL) [37] and kernel ridge regression confidence machine [38], are used for identifying the marine dynamic model. Among them, the GP have stronger robustness and generalization with a priori introduction from Bayesian perspective than other methods.…”

Section: Introductionmentioning

confidence: 99%

“…With regards to parametric gray-box modeling, ship dynamic models based on conjugate and semiconjugate Bayesian regression (ScBR) are adopted to estimate the hydrodynamic parameters in our previous work [44]. For the black-box modeling, Ariza Ramirez et al [36] used multioutput GPs to identify the ship dynamic system and showed that the GP scheme has better generalization than recurrent neural network (RNN). A series of Bayesian methods is used to quantify the extremal responses of a floating production storage and offloading (FPSO) vessel in [45].…”

Section: Introductionmentioning

confidence: 99%

Dynamic Model Identification of Ships and Wave Energy Converters Based on Semi-Conjugate Linear Regression and Noisy Input Gaussian Process

Liu

Xue

Huang

et al. 2021

JMSE

View full text Add to dashboard Cite

Reducing the carbon emissions of ships and increasing the utilization of marine renewable energy are the important ways to achieve the goal of carbon neutrality in ocean engineering. Establishing an accurate mathematical model is the foundation of simulating the motion of marine vehicles and structures, and it is the basis of operation energy efficiency optimization and prediction of power generation. System identification from observed input–output data is a practical and powerful method. However, for modeling objects with different characteristics and known information, a single modeling framework can hardly meet the requirements of model establishment. Moreover, there are some challenges in system identification, such as parameter drift and overfitting. In this work, three robust methods are proposed for generating ocean hydrodynamic models based on Bayesian regression. Two Bayesian techniques, semi-conjugate linear regression and noisy input Gaussian process regression are used for parametric and nonparametric gray-box modeling and black-box modeling. The experimental free-running tests of the KRISO very large crude oil carrier (KVLCC2) ship model and a multi-freedom wave energy converter (WEC) are used to validate the proposed Bayesian models. The results demonstrate that the proposed schemes for system identification of the ship and WEC have good generalization ability and robustness. Finally, the developed modeling methods are evaluated considering the aspects required conditions, operating characteristics, and prediction accuracy.

show abstract

“…Before designing the controller, it is necessary to identify the system dynamics. This process requires using an excitation and storing the input and output data, in order to obtain a model that can correctly describe the system behavior [2]. In this context, identification techniques arise, which use statistical methods to design mathematical models from experimental data of a process [3,4].…”

Section: Introductionmentioning

confidence: 99%

Evaluation of non-parametric identification techniques in second order models plus dead time

Robles-Algarín

Rodríguez

Ospino

2020

IJECE

View full text Add to dashboard Cite

In this paper, a set of non-parametric identification techniques are used in order to obtain second order models plus dead time for an underdamped system. Initially, non-parametric techniques were used to identify the system from the temperature data of a coal-heated oven. In this case, the identification techniques proposed by Stark, Jahanmiri - Fallahi and Ogata were used, which require obtaining two or three points of the step response for the system under study. In addition, the Matlab PID Tuner app was used to identify the underdamped system and compare the results with the other methods. The results show that the PID Tuner and the method proposed by Ogata are the ones that best represent the dynamics of the underdamped system, taking into account the values for the Integral Absolute Error (IAE) and the correlation coefficient. With the Stark method an IAE of 181.56 was obtained, while with the PID Tuner the best performance was achieved with an IAE of 21.59. In terms of the results obtained with the cross correlation, the best performance was achieved with the PID tuner and the Stark method.

show abstract

Non-parametric dynamic system identification of ships using multi-output Gaussian Processes

Cited by 42 publications

References 30 publications

Virtual environment for the design of position trajectory tracking controllers of remotely operated vehicles

Virtual environment for the design of position trajectory tracking controllers of remotely operated vehicles

Dynamic Model Identification of Ships and Wave Energy Converters Based on Semi-Conjugate Linear Regression and Noisy Input Gaussian Process

Evaluation of non-parametric identification techniques in second order models plus dead time

Contact Info

Product

Resources

About