Data-driven Modeling of Ship Motion Prediction Based on Support Vector Regression

Kawan, Bikram; Wang, Hao; Li, Guoyuan; Chhantyal, Khim

doi:10.3384/ecp17138350

Cited by 13 publications

(10 citation statements)

References 16 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thanks to the development of computational processing power, data-driven predictors have been gaining increasing attention. Previous research employed NNs [6], Support Vector Machines [15], Gaussian Process Regressions [16], and Auto-Regressive models [17]. The biggest advantage of a datadriven predictor is that it is able to deal with non-linearity and uncertainty through offline and online experiences.…”

Section: B Data-driven Predictionmentioning

confidence: 99%

A Multiple-Output Hybrid Ship Trajectory Predictor With Consideration for Future Command Assumption

Kanazawa

Skulstad

et al. 2021

IEEE Sensors J.

Self Cite

View full text Add to dashboard Cite

Onboard sensors contribute to data-driven understanding of complex and nonlinear ship dynamics in real time. By using sensors, precise ship trajectory prediction plays a key role in intelligent collision avoidance. A hybrid predictor makes prediction based on a mathematical model of which error is compensated by a black-box model. A Multiple-output Hybrid Predictor (MHP), which makes a long-horizon prediction at a time based on onboard sensor data, was developed in the previous study. However, it can not handle a time series of future command assumption. This limitation hinders an MHP from being applied to the evaluation of future command assumption in the predictive decision making. A novel architecture of MHP presented in this study converts a long time series of future command assumption into a fixed-length model-based-predicted vessel state; then, it is included in inputs of a black-box error compensator. This idea is robust to multidimensionality of commands and long control horizon. Assuming a low-fidelity vessel model and a limited data are available, simulation experiments are conducted. The effect of environmental disturbances and maneuverings on the prediction performance is examined comprehensively for the first time. The present study successfully incorporates a long time series of future command assumption in an MHP. It reduces the mean error by 81.8% compared to a model-based predictor and by 45.6% compared to a data-driven predictor under Beaufort wind force scale 4 wave, wind, and ocean current. Present study expands the application of MHP to the predictive decision making of future autonomous ships.

show abstract

Section: B Data-driven Predictionmentioning

confidence: 99%

A Multiple-Output Hybrid Ship Trajectory Predictor With Consideration for Future Command Assumption

Kanazawa

Skulstad

et al. 2021

IEEE Sensors J.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Both PSSP and TSSP models assume that the distance between the target trajectory and the trajectory most similar to it remains constant over the prediction duration. However, vessel movement is affected by external contexts, such as wind, waves and sea currents (Kawan et al, 2017;Sharif and Alesheikh, 2018;Kaffash-Charandabi et al, 2019). In addition, the maritime navigation environment is a complex system that varies over time (Tang et al, 2019).…”

Section: Point-based Similarity Search Prediction (Pssp)mentioning

confidence: 99%

“…The modeldriven models require that process parameters be estimated, which is not straightforward (Dalsnes et al, 2018). Besides, these models become complicated and need more time when non-linearity of the movement increases (Kawan et al, 2017). Data-driven models not only address the drawbacks of model-driven models, but also predict vessel trajectory for longer durations, which is required in collision avoidance.…”

mentioning

confidence: 99%

Vessel Trajectory Prediction Using Historical Automatic Identification System Data

2020

View full text Add to dashboard Cite

For maritime safety and security, vessels should be able to predict the trajectories of nearby vessels to avoid collision. This research proposes three novel models based on similarity search of trajectories that predict vessels' trajectories in the short and long term. The first and second prediction models are, respectively, point-based and trajectory-based models that consider constant distances between target and sample trajectories. The third prediction model is a trajectory-based model that exploits a long short-term memory approach to measure the dynamic distance between target and sample trajectories. To evaluate the performance of the proposed models, they are applied to a real automatic identification system (AIS) vessel dataset in the Strait of Georgia, USA. The models' accuracies in terms of Haversine distance between the predicted and actual positions show relative prediction error reductions of 40·85% for the second model compared with the first model and 23% for the third model compared with the second model.

show abstract

“…Our goal is to compute sensitivity index of every input parameter to construct a compact data-driven model. In this paper, SVM is selected as a tool to construct a surrogate regression model because of its high generalization, a better solution of non-linearity and high-dimensionality, avoidance of local optimal solution in ANN, which is explained in detail by [14]. For m training samples (x i , y i ) where x i ∈ R k and y i ∈ R The basic form of SVM is as follows:…”

Section: B Modelingmentioning

confidence: 99%

A SVM-based Sensitivity Analysis Approach for Data-Driven Modeling of Ship Motion

Wang

Cheng

Chen

et al. 2018

2018 IEEE International Conference on Mechatronics and Automation (ICMA)

View full text Add to dashboard Cite

This paper presents a novel method that combines support vector machine (SVM) with sensitivity analysis (SA) to analyze sensor data for ship motion modeling. In order to investigate how each model input contributes to the model's output, the PAWN method which is based on cumulative distribution function (CDF) is used. Considering the limitation of the PAWN method that it cannot be applied to sensor data directly, a surrogate model using SVM is integrated into PAWN method as it has a better solution of non-linearity and high-dimension problem as well as excellent generalization capability. Implementation of whole systematic workflow is elaborated in detail and related experiments are made such as comparison of four distance metric methods, benchmark test and SA of vessel data. The results show the proposed method can be used for analyzing vessel data to seek prime parameters that affect ship motion.

show abstract

Data-driven Modeling of Ship Motion Prediction Based on Support Vector Regression

Cited by 13 publications

References 16 publications

A Multiple-Output Hybrid Ship Trajectory Predictor With Consideration for Future Command Assumption

A Multiple-Output Hybrid Ship Trajectory Predictor With Consideration for Future Command Assumption

Vessel Trajectory Prediction Using Historical Automatic Identification System Data

A SVM-based Sensitivity Analysis Approach for Data-Driven Modeling of Ship Motion

Contact Info

Product

Resources

About