A study on vessel fatigue damage as a criterion for heading selection by application of 2D actual bimodal and JONSWAP wave spectra

Abstract: Planning and execution of marine operations requires proper estimation of vessel dynamic responses and their corresponding operational limits, including considerations of fatigue damage. Current guidelines for marine operations are based on a design wave height without considering the wave energy distribution in frequency and direction. This can be critical for ships operating in open seas where multimodal wave spectra may occur frequently. This study provides criteria for heading selection, with the aim of reducing fatigue damage of vessels under action of directional (2D) bimodal and multimodal wave spectra. In addition, some consequences of using analytical 2D JONSWAP spectra are also addressed. Based on a hydrodynamic model of a vessel, stresses at the midships section are computed using a spectral method. For bimodal wave spectra and considering that all dynamic responses are acceptable, fatigue damage can be reduced in about 50% when the vessel is heading to the least energetic wave component of 2D wave spectra. Moreover, fatigue damage obtained from actual 2D bimodal spectra can be well represented by its corresponding JONSWAP counterpart computed from the spectral parameters of the largest wave component. These findings can be used for vessel heading selection during planning and execution of marine operations.

Leer

A simulator of Synthetic Aperture Radar (SAR) image spectra: the applications on oceanswell waves

Abstract: The Synthetic Aperture Radar (SAR) carried on-board satellites yields invaluable data of global wave spectra since the early 1990s, with several satellites in orbit at present and more launches scheduled in the near future. However, the retrieval of wave information from SAR images constitutes a complex set of procedures. In this context, we have presented here a methodology to simulate SAR image spectra of ocean swell waves. SAR simulators are important tools for the implementation and evaluation of wave spectra retrieval schemes. The one proposed here is based on the Hasselmann Transform whose Modulation Transfer Functions (MTF’s) account for the main physical processes involved in the imaging of ocean waves. A detailed description of its structure is provided. Through several test cases, we highlight some particularities of the relationship between SAR image spectra and wave spectra. We have evaluated the impact of the parameters settings and input information on the retrieval process, pinpointing possible shortcomings. The results indicated that useful information about the processes involved in the imaging of ocean swells can be derived from the SAR simulator.

Leer

Learning the Sampling Pattern for MRI

Abstract: The discovery of the theory of compressed sensingbrought the realisation that many inverse problems can be solvedeven when measurements are "incomplete". This is particularlyinteresting in magnetic resonance imaging (MRI), where longacquisition times can limit its use. In this work, we considerthe problem of learning a sparse sampling pattern that can beused to optimally balance acquisition time versus quality of thereconstructed image. We use a supervised learning approach,making the assumption that our training data is representativeenough of new data acquisitions. We demonstrate that this isindeed the case, even if the training data consists of just 7training pairs of measurements and ground-truth images; with atraining set of brain images of size 192 by 192, for instance, oneof the learned patterns samples only 35% of k-space, howeverresults in reconstructions with mean SSIM 0.914 on a test setof similar images. The proposed framework is general enoughto learn arbitrary sampling patterns, including common patternssuch as Cartesian, spiral and radial sampling.

Leer

Predictive Online Optimisation with Applications to Optical Flow

Abstract: Online optimisation revolves around new data being introduced into a problem while it is still being solved; think of deep learning as more training samples become available. We adapt the idea to dynamic inverse problems such as video processing with optical flow. We introduce a corresponding predictive online primal-dual proximal splitting method. The video frames now exactly correspond to the algorithm iterations. A user-prescribed predictor describes the evolution of the primal variable. To prove convergence we need a predictor for the dual variable based on (proximal) gradient flow. This affects the model that the method asymptotically minimises. We show that for inverse problems the effect is, essentially, to construct a new dynamic regulariser based on infimal convolution of the static regularisers with the temporal coupling. We finish by demonstrating excellent real-time performance of our method in computational image stabilisation and convergence in terms of regularisation theory.

Leer