Prediction Of Roll Motion Research Articles

Premonition-driven deep learning model for short-term ship violent roll motion prediction based on the hull attitude premonition mechanism

Ship transit forecasting can provide advanced movement information to a multidimensional wave motion compensation system. However, traditional deep learning models based on data-driven or physical data have difficulty in accurately predicting the rapid variations in ship motion attitudes that frequently occur in complex ocean environments. Therefore, a premonition-driven deep learning model is proposed in this paper to significantly improve the prediction accuracy of ship violent roll motion. This model applies precursor information to predict premonitory values of ship prospective motion attitudes based on an integrated statistical regression model and then employs these premonitory values to revise the outputs of the deep learning time series model. First, a hull attitude premonition mechanism is presented based on biological premonition cognitive behaviour, which captures the correlation relationships between extreme values by downgrading the globally continuous time series prediction problem to a time-independent regression problem. The corresponding methods for predicting local extreme values and deep feedforward neural network cluster are established based on these correlation relationships between the extreme values to achieve precise premonitory values. Second, the time and amplitude variables of the extreme angle values are transformed to countdown time variables and sequential sets of premonitory angle values. Two types of premonition units and their respective premonitory long short-term memory neural networks are constructed to upgrade the regression problem to a globally continuous time series prediction problem. Finally, a single-step output recurrent strategy is applied to forecast the short-term ship motion attitude for fifteen seconds by considering eleven initially predicted points across the two datasets. In sixty-six simulation experiments, the mean absolute error, root mean squared error, and mean absolute percentage error of the premonition-driven long short-term memory neural networks are considerably lower than those of traditional deep learning models. These simulation experiments verify that the premonition-driven deep learning model can effectively forecast short-term violent ship roll motion and provide safeguards for cargo transfer.

Roll Prediction and Parameter Identification of Marine Vessels Under Unknown Ocean Disturbances

Abstract This paper deals with two topics: roll predictions of marine vessels with machine-learning methods and parameter estimation of unknown ocean disturbances when the amplitude, frequency, offset, and phase are difficult to estimate. This paper aims to prevent the risky roll motions of marine vessels exposed to harsh circumstances. First of all, this study demonstrates complex dynamic phenomena by utilising a bifurcation diagram, Lyapunov exponents, and a Poincare section. Without any observers, an adaptive identification applies these four parameters to the globally exponential convergence using linear second-order filters and parameter estimation errors. Then, a backstepping controller is employed to make an exponential convergence of the state variables to zero. Finally, this work presents the prediction of roll motion using reservoir computing (RC). As a result, the RC process shows good performance for chaotic time series prediction in future states. Thus, the poor predictability of Lyapunov exponents may be overcome to a certain extent, with the help of machine learning. Numerical simulations validate the dynamic behaviour and the efficacy of the proposed scheme.

Short-term ship roll motion prediction using the encoder–decoder Bi-LSTM with teacher forcing

The safety of maritime operations has become a paramount concern with the advancement of intelligent ships. Ship stability and safety are directly impacted by roll motion, making the prediction of short-term ship roll motion pivotal for assisting navigators in making timely adjustments and averting hazardous roll conditions. However, predicting ship roll motion poses challenges due to nonlinear dynamics. This study aims to predict short-term ship roll motion by leveraging the encoder–decoder structure of Bidirectional Long Short-Term Memory Networks (Bi-LSTM) with teacher forcing. The model is accomplished by employing an encoder–decoder structure to map input sequences to output sequences of varying lengths, and employing teacher forcing to enhance the network’s ability to extract information. To refine and analyze the prediction model, aspects such as the quantity of training data to guarantee model generalization, establishing apposite length relationships between input and output sequences, and assessing the model performance in various sea states are investigated. Additionally, comparative experiments assessing roll motion prediction for intervals of 10s, 30 s, 60 s, and 120 s are conducted to substantiate the necessity and effectiveness of the proposed network. The dataset originates from a commercial professional simulator developed by the Norwegian company Offshore Simulator Center AS (OSC).

Probabilistic analysis of the instantaneous center of rotation of a floating structure under regular and irregular beam waves

Accurate prediction of roll motion is crucial to the safety of ships and floating structures. The “roll center” mentioned in the literature is ambiguous and various definitions exist. Hence, the instantaneous center of rotation (ICR) is preferred, being an established kinematic concept. Understanding the ICR is important on account of its significant influence on roll damping. A previous study postulated that the ICR moves in a straight line for regular beam waves, based on experimental observations. In this paper, the probability distribution of the ICR coordinates (y1, y2) is derived analytically for regular and irregular beam waves. For regular waves, it is proven a fortiori that the ICR lies on a straight line, and that y1 and y2 follow the Cauchy distribution, for which the mean and variance are undefined. For irregular waves, it is shown that the joint probability density of y1 and y2 is a bivariate Cauchy distribution, thus the iso-density contour lines are concentric ellipses. Regular and irregular wave experiments are performed to investigate realistic conditions. The analytical solutions are found to be quite accurate, notwithstanding some discrepancies arising from nonlinearities and experimental imperfections. The theoretical results presented herein have potential practical applications, for example since the ICR governs damping, the ICR distribution allows damping to be modeled as a stochastic process.

Formulation of ship roll damping models from free-decay data

Estimation of roll damping of a ship is essential to the prediction of its response in waves. The formulation of ship roll damping models from experiment or CFD data involves two parts: the determination of the model form and the calculation of the model coefficients, and both have significant influences on damping accuracy. To improve the roll motion prediction of a particular ship, this paper made an investigation of the typical damping models and its influence factors based on benchmark experiment data. The results show that the typical damping models can not reflect the fluid physics of roll motion. A conclusion is given that all roll angle, angular roll velocity, and angular roll acceleration should be considered in damping model determination. For the convenience of solving the roll motion equation, damping models with angle and angular velocity as independent variables are proposed due to the coupling of the angle, angular velocity, and the angular acceleration in roll. A procedure of damping model formulation is developed based on the Prony-SS method which was introduced into roll damping identification recently. Examples that follow the procedure show good performance. Furthermore, forward speed and flow memory effects on ship roll damping are discussed.

Numerical study of the interaction between the pure loss of stability, surf-riding, and broaching on ship capsizing

Ship capsizing in stern quartering wave is very complicate with a strong correlation to the stability failure modes proposed in second-generation intact stability criteria. It is of great importance to shed some light on this complex phenomenon. In this study, the capsizing behavior of a free-running ONR Tumblehome model in stern quartering seas is investigated by an unsteady Reynolds-averaged Navier–Stokes (URANS) solver coupled with dynamic overset grid approach. Validation study of the URANS solver is performed for the prediction of extreme roll motion of the ship. The comparisons show satisfactory agreement, with an average difference less than 3% between the simulation and experimental results. The correlation between the occurrence of stability failure modes (surf-riding, pure loss of stability, and broaching) and ship capsizing is then assessed under different ship speed and heading angle conditions. It is concluded that the capsizing occurs as the free-running ship sailing with a high speed, which is caused by the interaction between the broaching and severe stability loss. In addition, influenced by the wave-induced moment, both the pure loss of stability and surf-riding modes may cause the broaching, indicating the similar occurring condition and strong relation between these stability failure modes.

Short-time FORM analysis for extreme roll motion prediction in beam seas

This paper presents an efficient computational method to predict extreme roll motions by the First Order Reliability Method (FORM). The method facilitates accurate estimation of the reliability index distribution by short time domain calculations. The calculations are based on the transfer function of roll motion together with an equivalent linear damping for the estimation of the most probable largest roll motion using nonlinear time domain simulations. The linearization of the damping gives rise to an inconsistency (a “time shift”) in the time series evaluation. Thus, in order to account for this nonlinear effect, when evaluating the most probable largest roll motions, a method to correct the time shift from the linear solution is proposed. Two different container ships in beam seas and zero forward speed are used to demonstrate the effectiveness of the method. It is also shown that a further reduction of computational efforts is attained by using the Karhunen-Loeve wave representation based on Prolate Spheroidal Wave Functions (PSWF).

A Study on the Number of Preferable Neighborhood Vectors When Using the Improved Jacobi Matrix Estimation Method for Roll Motion Prediction of Small Boat

A Study on the Number of Preferable Neighborhood Vectors When Using the Improved Jacobi Matrix Estimation Method for Roll Motion Prediction of Small Boat

Roll Motion Prediction of Unmanned Surface Vehicle Based on Coupled CNN and LSTM

The prediction of roll motion in unmanned surface vehicles (USVs) is vital for marine safety and the efficiency of USV operations. However, the USV roll motion at sea is a complex time-varying nonlinear and non-stationary dynamic system, which varies with time-varying environmental disturbances as well as various sailing conditions. The conventional methods have the disadvantages of low accuracy, poor robustness, and insufficient practical application ability. The rise of deep learning provides new opportunities for USV motion modeling and prediction. In this paper, a data-driven neural network model is constructed by combining a convolution neural network (CNN) with long short-term memory (LSTM) for USV roll motion prediction. The CNN is used to extract spatially relevant and local time series features of the USV sensor data. The LSTM layer is exploited to reflect the long-term movement process of the USV and predict roll motion for the next moment. The fully connected layer is utilized to decode the LSTM output and calculate the final prediction results. The effectiveness of the proposed model was proved using USV roll motion prediction experiments based on two case studies from “JingHai-VI” and “JingHai-III” USVS of Shanghai University. Experimental results on a real data set indicated that our proposed model obviously outperformed the state-of-the-art methods.

Study on Roll Restoring Arm Variation Using a Three-Dimensional Hybrid Panel Method

Providing a reliable numerical tool for the prediction of parametric roll and pure loss of stability is a difficult task. It is acknowledged that the occurrence of these two failure modes is mainly attributed to the variation of roll restoring arm, which should be calculated accurately. This work studies exclusively the roll restoring arm variation numerically and experimentally. First, partially restrained experimental tests at different constant heeling angles (up to 30°) are conducted to measure the variation of roll restoring arm in regular waves. Second, a numerical method for the calculation of roll restoring arm in waves taking radiation and diffraction moments into account is adopted. The calculation results are benchmarked against experimental model test data. Third, the characteristics of roll restoring arm variation and the effects of radiation and diffraction moments on roll restoring arm variation (GZRD) in head and following seas are investigated. Finally, the feasibility of GZ calculated by the integral of instantaneously average wetted surface is studied by comparing the results with that calculated by the integral of the fully instantaneously wetted surface. The research show that the GZRD can be ignored in following seas. The GZ in head seas consists of two important harmonic components, and GZRD is the main contribution of the second harmonic component. Therefore, GZRD should be calculated with consideration of the instantaneous heeling angles and Froude number in head seas, especially for the prediction of large amplitude roll motion at high speed. 1. Introduction Currently, the second-generation intact stability criteria are under development by the International Maritime Organization (IMO) (IMO SLF 53/WP.1-Add.1 2011; IMO SDC 3/WP.5 2016). Five different kinds of potentially dangerous dynamic stability failure modes in waves were proposed. Parametric roll and pure loss of stability are two of the stability failure modes that may lead to capsizing of a ship. According to the proposed drafts, a multilevel structure was adopted to make sure that numerical assessment procedures are only applied when the vulnerability to dynamical stability failures is beyond reasonable doubt. Assessment of the vulnerability criteria is normally based on simplified geometry characteristics or empirical formulas. However, direct stability assessment requires highly accurate prediction (Kobylinski 2012).

Ship Roll Damping Coefficient Prediction Using CFD

For a ship undergoing roll motion at a resonance frequency, the roll restoring moment and virtual moment of inertia counteract each other and only the damping moment resists the roll exciting moments. Hence, accurate computation of the roll damping moment improves the roll motion prediction. There are several methods to calculate the damping moment; however, most of them fail to consider the viscous effects, such as the potential theory-based method. The experimental measurements take into account the viscous effect; however, they are expensive. The Computational Fluid Dynamics (CFD) method is an alternative to accurately consider the viscous effect. In this study, a CFD approach based on harmonic exciting roll motion technique is adopted to compute the roll motion characteristics and damping coefficients at different conditions. The impact of appendages, Froude number, and degrees of freedom on roll motion characteristics and damping coefficients are investigated for the model scale and full scale to study the scale effects. The results at the model scale are compared against experimental measurements and good correlation was found. 1. Introduction Controlling the roll motion of a ship in a seaway is essential for the safety and habitability. The roll motion of a ship is nonlinear, and the nonlinearity increases at the resonance as the ship experiences larger roll motion. Accurately computing the roll damping at the resonance frequency is important because the moment of inertia and restoring moment counteract each other and the damping moment opposes the roll motion. The total roll damping consists of friction damping, body eddy shedding damping, surface wave damping, lift force damping, and appendages damping. The wave and lift force damping can be calculated by potential flow theory based methods, whereas others are influenced by viscosity. In general, most roll damping calculation methods are based on empirical models, and Ikeda's method (Ikeda et al. 1978) is most common one. Although this method works quite well for conventional ships, the predicted results are sometimes conservative or underestimated for unconventional ships (Gu et al. 2013). Large roll damping is strongly nonlinear and is influenced by fluid viscosity and flow characteristics such as flow separation and vortex shedding. Empirical or semi-empirical methods cannot account the flow characteristics of such a complex flow. Roll damping calculation for conventional ships by Ikeda's simplified method can fit experimental data quite well at small roll angles. However, when the roll angle is large, it is out of the acceptable range of Ikeda's method, and the accuracy of the computed damping coefficient is low (Kawahara et al. 2012).

Prediction of a ship roll added mass moment of inertia using numerical simulation

Prediction of a ship's roll motion in parametric roll and dead ship condition using CFD and experiments is time consuming and expensive, while a reliable equation based method can drastically decrease the computation time. The commonly used roll motion equation contains roll mass and added mass moment of inertia, damping and restoring terms and associated coefficients. Improving the prediction of these coefficients increases the accuracy of the method. There is a lack of precise calculation for the roll added mass moment of inertia, which is commonly estimated using semi-empirical techniques and as a percentage of the roll mass moment of inertia coefficient. This paper presents a new method to determine the roll added mass moment of inertia using CFD simulations based on a harmonic excitation roll motion technique and the results are compared against Bhattacharyya's method. The influences of excitation frequency, Froude number, roll angle and appendages on the roll motion characteristics and the roll added mass moment of inertia coefficient are investigated. In addition, the roll added mass moment of inertia coefficients on the model scale and full scale ship are computed to investigate the scale effect.

Ship roll motion prediction based on ℓ1 regularized extreme learning machine.

In this paper, a new method is proposed for prediction of ship roll motion based on extreme learning machine (ELM). To improve the prediction accuracy and avoid over or under fitting, two techniques are adopted to select the appropriate structure of ELM. First, the inputs of the ELM are selected from the roll motion time series using Lipschitz quotient method. Second, the number of hidden layer nodes is determined via ℓ1 regularized technique. Finally, the ℓ1 regularized ELM is solved by least angle regression (LAR) algorithm. The effectiveness of the proposed method is demonstrated by ship roll motion prediction experiments based on the real measured ship roll motion time series.

Numerical Assessment of Roll Motion Characteristics and Damping Coefficient of a Ship

Accurate calculation of the roll damping moment at resonance condition is essential for roll motion prediction. Because at the resonance condition, the moment of inertia counteracts restoring moment and only the damping moment resists increase in the roll angle. There are various methods to calculate the roll damping moment which are based on potential flow theory. These methods have limitations to taking into account the viscous effects in estimating the roll motion, while, CFD as a numerical method is capable of considering the viscous effects. In this study, a CFD method based on a harmonic excited roll motion (HERM) technique is used to compute the roll motion and the roll damping moment of a containership’s model in different conditions. The influence of excitation frequency, forward speed and degrees of freedom at beam-sea and oblique-sea realizations are considered in estimating the roll damping coefficients. The results are validated against model tests, where a good agreement is found.

A real-time ship roll motion prediction using wavelet transform and variable RBF network

Real-time prediction of ship roll motion is vital for marine safety and efficiency of operations onboard the ship. However, ship roll motion is a complex time-varying nonlinear process which varies with various sailing conditions as well as time-varying environmental factors. To achieve precise real-time ship roll prediction, an ensemble prediction scheme is constructed by combining the discrete wavelet transform (DWT) method with the variable-structure radial basis function (RBF) network. The DWT is used to reduce the time-series data redundancies and carry the data information in few significant uncoupled sub-series, thus facilitate the identification and prediction by using the variable RBF networks. The variable RBF networks are used to represent time-varying dynamics with both the structure and parameters are tuned in real time. The DWT-transform-based variable RBF networks are used to represent the time-varying nonlinear dynamics of ship roll movement during ship maneuvering. The effectiveness of the proposed DWT-based real-time roll prediction scheme is demonstrated by short-term ship roll motion prediction experiments based on the actual ship roll motion measurements collected during sea test of M.V. YuKun.

Roll motion prediction using a hybrid deep learning and ARIMA model

Autoregressive Integrated Moving Average (ARIMA) is one of the linear model that is good, flexible, and easy to use in univariate time series analysis and forecasting. Some research activities in time series forecasting also suggest Artificial Neural Network (ANN) model as an alternative nonlinear model for forecasting. ARIMA model has a good ability to capture the linear pattern while the ANN model is good to capture the nonlinear pattern. ARIMA and ANN models have been widely used in the prediction of roll motion. ARIMA and ANN can also be combine as a hybrid model to take advantage of the ability of ARIMA and ANN models in linear and nonlinear modeling compared to ARIMA and ANN model. In this paper, we adapt the hybrid methodology to combine ARIMA and Deep Neural Network (DNN) model, an ANN model with multiple hidden layers. The real dataset used is the roll motion of a Floating Production Unit (FPU). The empirical results show that the DNN-ARIMA hybrid model is the best model for predicting the roll motion compared to the non hybrid models and very effective to improve forecast accuracy.

Query-specific optimal convolutional neural ranker

In this paper, we propose a novel learning-to-rank method by developing a convolutional neural network (CNN)-based ranking score estimation function (ranker). We use the query, query-specific preference, and the neighborhood structure to regularize the learning of the CNN ranker parameters. We propose to impose the CNN outputs of a query-preferred data object to be larger than that of a data object which the query tries to avoid. Also we hope the ranking scores of the data objects can be smooth over neighborhoods and the ranking score of the query itself can be large. We construct a joint unified minimization problem by combining these regularization problems to learn the parameters of CNN, and develop an iterative algorithm based on fix-point method. The experiments over the benchmark data sets of image retrieval and ship roll motion prediction show its effectiveness.

Application of advanced system identification technique to extract roll damping from model tests in order to accurately predict roll motions

In this paper our previously developed advanced system identification technique [1] has been applied to extract the frequency dependent roll damping from a series of model tests run in irregular (random) waves. It is shown that this methodology accurately models the roll damping which can then be used to produce accurate predictions of the ships roll motion. These roll motion predictions are not only more accurate than the potential flow predictions but more accurate than potential flow models corrected using either empirical prediction methods [2] and even those corrected using roll damping obtained from free decay sallying experiments. This methodology has the potential to significantly improve roll motion prediction during full scale at sea trails of vessels in order to dramatically improve safety of critical operations such as helicopter landing or ship to ship cargo transfer.

Bilge-Keel Influence on Free Decay of Roll Motion of a Realistic Hull1

The prediction of roll motion of a ship with bilge keels is particularly difficult because of the nonlinear characteristics of the viscous roll damping. Flow separation and vortex shedding caused by bilge keels significantly affect the roll damping and hence the magnitude of the roll response. To predict the ship motion, the Slender-Ship Free-Surface Random-Vortex Method (SSFSRVM) was employed. It is a fast discrete-vortex free-surface viscous-flow solver developed to run on a standard desktop computer. It features a quasi-three-dimensional formulation that allows the decomposition of the three-dimensional ship-hull problem into a series of two-dimensional computational planes, in which the two-dimensional free-surface Navier–Stokes solver Free-Surface Random-Vortex Method (FSRVM) can be applied. In this paper, the effectiveness of SSFSRVM modeling is examined by comparing the time histories of free roll-decay motion resulting from simulations and from experimental measurements. Furthermore, the detailed two-dimensional vorticity distribution near a bilge keel obtained from the numerical model will also be compared with the existing experimental Digital Particle Image Velocimetry (DPIV) images. Next, we will report, based on the time-domain simulation of the coupled hull and fluid motion, how the roll-decay coefficients and the flow field are altered by the span of the bilge keels. Plots of vorticity contour and vorticity isosurface along the three-dimensional hull will be presented to reveal the motion of fluid particles and vortex filaments near the keels.

Development of a semi-analytical method for the efficient prediction of parametric roll

A semi-analytical approach to the simulation of the parametric roll motions of container ships in head seas is presented. In the proposed 1.5-degree of freedom model, the nonlinear change in the transverse stability induced by the vertical motions and wave elevation is approximated by investigating the relationship between the restoring lever (GZ) and the metacentric height (GM) in waves. In particular, the variations in the GM in irregular waves are modeled by the transfer functions of the mean value and first harmonic component on the basis of the assumption of superposition. Using the derived formula, the quantitative prediction of parametric roll motion as well as the susceptibility to such motion in regular and irregular waves is conducted. A comparison with the results of a weakly nonlinear time-domain method, namely, the impulse response function method based on the hydrodynamic coefficients of strip theory, is carried out to validate the proposed model. Not only the accuracy and efficiency of the present approach but also the stochastic properties of parametric roll in random seas are examined considering the sensitivities and uncertainties in the computations.