Time-Frequency Collaborative Learning for Imbalanced Ship Motion Data With Missing Labels in Sea State Estimation.

Very short-term ship motion forecasting aims to predict future movements using historical ship motion data. While ship motion features both temporal and frequency characteristics, the latter is often neglected. This paper proposes a fully adaptive time–frequency coupling forecasting model using self-attention mechanism based on the Deep Operator Network (DeepONet), abbreviated as TFD. The multi-head attention layers enable the trunk net to adaptively learn the relationships between different frequencies in the frequency domain and assign varying weights accordingly. Thus, compared to the Fourier transform and multilayer perceptron-net enhance model based on DeepONet (FMD), which relies on manually specified filter frequencies, the TFD model is capable of fully learning the motion patterns in both the time and frequency domains, establishing nonlinear mapping relationships between them. It exhibits greater interpretability and generalization. The TFD model is tested for accuracy and generalization using ship motion data from the Iowa University experimental tank. The results indicate that, compared to the DeepONet and FMD, the TFD model reduces the mean square error (MSE) by up to 64.72% and 52.45%, with an average reduction of 55.57% and 42.47%. In terms of generalization, the forecasting MSE is reduced by up to 65.04% and 46.08%. Compared to the DeepONet and FMD, the proposed TFD model demonstrates significant improvements in forecasting horizon and generalization, providing a notable advantage in very short-term ship motion prediction applications.

Accurate prediction of ship motion is very important for ensuring marine safety, weapon control, and aircraft carrier landing, etc. Ship motion is a complex time-varying nonlinear process which is affected by many factors. Time series analysis method and many machine learning methods such as neural networks, support vector machines regression (SVR) have been widely used in ship motion predictions. However, these single models have certain limitations, so this paper adopts a multi-model prediction method. First, ensemble empirical mode decomposition (EEMD) is used to remove noise in ship motion data. Then the random forest (RF) prediction model optimized by genetic algorithm (GA), back propagation neural network (BPNN) prediction model and SVR prediction model are respectively established, and the final prediction results are obtained by results of three models. And the weights coefficients are determined by the correlation coefficients, reducing the risk of prediction and improving the reliability. The experimental results show that the proposed combined model EEMD-GARF-BPNN-SVR is superior to the single predictive model and more reliable. The mean absolute percentage error (MAPE) of the proposed model is 0.84%, but the results of the single models are greater than 1%.

Ship motions are usually employed to analyze water transportation system. Interpolation of ship motions can be used for estimating the lost ship movement information, which is important for analyzing water traffic. Although some methods are proposed to interpolate ship trajectory, these methods are needed to improve interpolation accuracy. In addition, the course and speed of ship are not interpolated in previous studies. This article proposes a novel estimation algorithm for interpolating ship motions. First, the bilateral filtering is used to smooth the navigation trajectories, then a novel interpolation approach forecasts the vessel positions during the interpolation time period by taking into account the dynamic information of Automatic Identification System (AIS), and then forms the interpolation track of the moving vessel. In order to validate the proposed method, the AIS data of Yangtze River is collected from the AIS stations and the proposed restoration methods are verified by MATLAB under simulation. The results show that the proposed method can effectively remove the inaccurate ship motion data and interpolate the lost ship motion data.

In this paper, the disturbance applied to launcher hatch by ship motions is introduced to identify the vertical ship motion disturbance. Basically, ship motions are comprised of 6 degrees of freedom: roll, pitch, yaw, heave, surge and sway. In the case of the shipboard launcher hatch the coupled pitch, heave and roll are significant motions to be transformed to a vertical direction motion. The maximum acceleration values are obtained from the vertical motion model and the ship motion data in accordance with ship type and hatch location on the ship. We verify that the maximum pitch motion mainly influences the launcher hatch and also present the quantity of the maximum load disturbance by the ship's motion acceleration.

The ship motions in waves are regarded as a nonstationary random process because of the ship maneuvers, such as changes of the course and the speed, even on the assumption that the seaway can be considered as stationary stochastic process. In this paper, the recursive type of Bayesian modeling procedure is proposed and applied to online estimation of directional wave spectra based on ship motion data, which are pitch angle, roll angle and vertical acceleration. In the proposed method, the convergence of the estimates is not achieved every time step, but gradually achieved with time steps by using the ship motion cross spectra. To cope with the nonstationarity included in the time history data, instantaneous cross spectrum analysis based on T-VVAR (Time Varying Coefficient Vector Auto Regressive) modeling procedure is adopted. The T-VVAR model is transformed into the state space model and the time varying coefficients can be evaluated by using the Kalman Filter algorithm. Using the estimated time varying coefficients, the instantaneous ship motion cross spectra can be calculated every moment. Moreover, in order to estimate the directional wave spectra stably, a new constraint condition, which is based on the observed significant wave height data, is introduced into the Bayesian modeling procedure. To examine the reliability of the proposed procedure, onboard tests were carried out on nonstationary conditions. It is shown that the statistical values derived by integration of the estimated directional wave spectra agree well with the observed one.

In-service monitoring can help to increase safety of ships especially regarding the fatigue assessment. For this purpose, it is compulsory to know the environmental conditions encountered: wind, but also the full directional wave spectrum. During the EU TULCS project, a full scale measurements campaign has been conducted onboard the CMA-CGM 13200 TEU container ship Rigoletto. She has been instrumented to measure deformation of the ship as well as the sea state encountered during its trip. This paper will focus on the sea state estimation. Three systems have been installed to estimate the sea state encountered by the Rigoletto: An X-band radar from Ocean Waves with WAMOS® system and two altimetric wave radars from RADAC®. Nevertheless, the measured significant wave height can be disturbed by several external elements like bow waves, sprays, sea surface ripples, etc… Furthermore, ship motions are also measured and can provide another estimation of the significant wave height using a specific algorithm developed by DCNS Research for the TULCS project. As all those estimations are inherently different, it is necessary to make a fusion of those data to provide a single estimation (“best estimate”) of the significant wave height. This paper will present the data fusion process developed for TULCS and show some first validation results.

Bayesian estimation of directional wave spectra based on ship motions

<i>The initial response to a marine accident can play a key role to minimize the accident. Therefore, various decision support systems have been developed using sensors, simulations, and active response equipment. In this study, we developed an algorithm to predict damage locations using ship motion data with bidirectional long short-term memory (BiLSTM), a type of recurrent neural network. To reflect the low frequency ship motion characteristics, 200 time-series data collected for 100 s were considered as input values. Heave, roll, and pitch were used as features for the prediction model. The F1-score of the BiLSTM model was 0.92; this was an improvement over the F1-score of 0.90 of a prior model. Furthermore, 53 of 75 locations of damage had an F1-score above 0.90. The model predicted the damage location with high accuracy, allowing for a quick initial response even if the ship did not have flood sensors. The model can be used as input data with high accuracy for a real-time progressive flooding simulator on board.</i>

Based on Bayesian modeling procedure, a stochastic procedure is introduced. Using this procedure, directional wave spectra can be estimated by using the ship motion data which were recorded during the ship underway. The prior distribution in this procedure consists of two parts. The one is a stochastic constraint so called Gaussian smoothness prior distribution. And the other one is adopted to avoid excessive estimation for the directional wave spectra. Furthermore, the encounter frequencies are transformed into the true wave frequencies even in following seas.In order to examine the reliability of the proposed procedure, the onboard tests were carried out. And, the possibility of this procedure can be shown and problems encountered on applying are reported.

: Beam steering angles of a shipboard phased array antenna are calculated, given the position of the ship, the position and attitude of the antenna on the ship, and the ship's bearing, roll, and pitch angles. The coordinate transformations were written in MATLAB; the pertinent software is added in the Appendix. An example of an application is given.

As software continues to insinuate itself into nearly every aspect of our life, the quality of software has been an extremely important issue. Software Quality Assurance (SQA) is a process that ensures the development of high-quality software. It concerns the important problem of maintaining, monitoring, and developing quality software. Accurate detection of fault prone components in software projects is one of the most commonly practiced techniques that offer the path to high quality products without excessive assurance expenditures. This type of quality modeling requires the availability of software modules with known fault content developed in similar environment. However, collection of fault data at module level, particularly in new projects, is expensive and time-consuming. Semi-supervised learning and active learning offer solutions to this problem for learning from limited labeled data by utilizing inexpensive unlabeled data.;In this dissertation, we investigate semi-supervised learning and active learning approaches in the software fault prediction problem. The role of base learner in semi-supervised learning is discussed using several state-of-the-art supervised learners. Our results showed that semi-supervised learning with appropriate base learner leads to better performance in fault proneness prediction compared to supervised learning. In addition, incorporating pre-processing technique prior to semi-supervised learning provides a promising direction to further improving the prediction performance. Active learning, sharing the similar idea as semi-supervised learning in utilizing unlabeled data, requires human efforts for labeling fault proneness in its learning process. Empirical results showed that active learning supplemented by dimensionality reduction technique performs better than the supervised learning on release-based data sets.

A semi-supervised deep learning approach for cropped image detection

Nonparametric modeling of ship maneuvering motion in waves based on Gaussian process regression

An analysis of the shipboard characteristics of the SH-2F helicopter in response to prescribed deck motion, deck friction, and steady wind conditions has been developed. The objective of deriving the SH-2F shipboard dynamic model is to define safe conditions for launching and recovering the helicopter from the flight deck of U.S. Navy frigates and destroyers. Operational conditions of interest include helicopter and ship deck dynamic interactions that could potentially cause dangerous interference between the helicopter and the ship, such as sliding or tipping of the helicopter. The wind condition, ship deck motion, helicopter rotor thrust, and friction coefficients between helicopter tires and flight-deck surfaces are found to be important parameters that affect the helicopter shipboard operations. Four sets of aerodynamic characteristics are modeled in the analysis: one with the rotor operating; one with the rotor stopped and inoperative; one with rotor folded; and one for the fuselage. The ship motion data, including three linear translation and two angular rotation degrees of freedom (roll and pitch) are described in the time domain. The equations of motion of the shipboard dynamic model are derived using the energy method.

In a majority–minority classification problem, class imbalance in the dataset(s) can dramatically skew the performance of classifiers, introducing a prediction bias for the majority class. Assuming the positive (minority) class is the group of interest and the given application domain dictates that a false negative is much costlier than a false positive, a negative (majority) class prediction bias could have adverse consequences. With big data, the mitigation of class imbalance poses an even greater challenge because of the varied and complex structure of the relatively much larger datasets. This paper provides a large survey of published studies within the last 8 years, focusing on high-class imbalance (i.e., a majority-to-minority class ratio between 100:1 and 10,000:1) in big data in order to assess the state-of-the-art in addressing adverse effects due to class imbalance. In this paper, two techniques are covered which include Data-Level (e.g., data sampling) and Algorithm-Level (e.g., cost-sensitive and hybrid/ensemble) Methods. Data sampling methods are popular in addressing class imbalance, with Random Over-Sampling methods generally showing better overall results. At the Algorithm-Level, there are some outstanding performers. Yet, in the published studies, there are inconsistent and conflicting results, coupled with a limited scope in evaluated techniques, indicating the need for more comprehensive, comparative studies.