A Review of SPH Techniques for Hydrodynamic Simulations of Ocean Energy Devices

This article is dedicated to providing a detailed review concerning the SPH-based hydrodynamic simulations for ocean energy devices (OEDs). Attention is particularly focused on three topics that are tightly related to the concerning field, covering (1) SPH-based numerical fluid tanks, (2) multi-physics SPH techniques towards simulating OEDs, and finally (3) computational efficiency and capacity. In addition, the striking challenges of the SPH method with respect to simulating OEDs are elaborated, and the future prospects of the SPH method for the concerning topics are also provided.

Deep learning for enhancing multi-source reverse time migration

<p>Reverse time migration (RTM) is a technique used to obtain high-resolution images of underground reflectors; however, this method is computationally intensive when dealing with large amounts of seismic data. Multi-source RTM can significantly reduce the computational cost by processing multiple shots simultaneously. However, multi-source-based methods frequently result in crosstalk artifacts in the migrated images, causing serious interference in the imaging signals. Plane-wave migration, as a mainstream multi-source method, can yield migrated images with plane waves in different angles by implementing phase encoding of the source and receiver wavefields; however, this method frequently requires a trade-off between computational efficiency and imaging quality. We propose a method based on deep learning for removing crosstalk artifacts and enhancing the image quality of plane-wave migration images. We designed a convolutional neural network that accepts an input of seven plane-wave images at different angles and outputs a clear and enhanced image. We built 505 1024×256 velocity models, and employed each of them using plane-wave migration to produce raw images at 0°, ±20°, ±40°, and ±60° as input of the network. Labels are high-resolution images computed from the corresponding reflectivity models by convolving with a Ricker wavelet. Random sub-images with a size of 512×128 were used for training the network. Numerical examples demonstrated the effectiveness of the trained network in crosstalk removal and imaging enhancement. The proposed method is superior to both the conventional RTM and plane-wave RTM (PWRTM) in imaging resolution. Moreover, the proposed method requires only seven migrations, significantly improving the computational efficiency. In the numerical examples, the processing time required by our method was approximately 1.6% and 10% of that required by RTM and PWRTM, respectively.</p>

Deep learning for enhancing multi-source reverse time migration

<p>Reverse time migration (RTM) is a technique used to obtain high-resolution images of underground reflectors; however, this method is computationally intensive when dealing with large amounts of seismic data. Multi-source RTM can significantly reduce the computational cost by processing multiple shots simultaneously. However, multi-source-based methods frequently result in crosstalk artifacts in the migrated images, causing serious interference in the imaging signals. Plane-wave migration, as a mainstream multi-source method, can yield migrated images with plane waves in different angles by implementing phase encoding of the source and receiver wavefields; however, this method frequently requires a trade-off between computational efficiency and imaging quality. We propose a method based on deep learning for removing crosstalk artifacts and enhancing the image quality of plane-wave migration images. We designed a convolutional neural network that accepts an input of seven plane-wave images at different angles and outputs a clear and enhanced image. We built 505 1024×256 velocity models, and employed each of them using plane-wave migration to produce raw images at 0°, ±20°, ±40°, and ±60° as input of the network. Labels are high-resolution images computed from the corresponding reflectivity models by convolving with a Ricker wavelet. Random sub-images with a size of 512×128 were used for training the network. Numerical examples demonstrated the effectiveness of the trained network in crosstalk removal and imaging enhancement. The proposed method is superior to both the conventional RTM and plane-wave RTM (PWRTM) in imaging resolution. Moreover, the proposed method requires only seven migrations, significantly improving the computational efficiency. In the numerical examples, the processing time required by our method was approximately 1.6% and 10% of that required by RTM and PWRTM, respectively.</p>

Ion Current Screening Modeling of the Ion–Weibel Instability

Ion–Weibel instability (IWI) is an important mechanism of generating a magnetic field in supernova remnants; it plays a key role in the generation of high-energy cosmic rays. Computational efficiency has been a bottleneck in numerical exploration of the large-scale evolution of IWI. Here I report a new hybrid particle-in-cell (PIC) method that can quickly simulate IWI. The method is based on a new model that describes the relation of the ion current and its magnetic field under the electron screening. The new method’s computational efficiency is nearly two orders of magnitude higher than that of the PIC method. This method is suitable for the full-scale simulation of the IWI in laser-plasma experiments and supernova remnants.

A Stereo Visual-Inertial Odometry Using Structural Lines for Localizing Indoor Wheeled Robots

This paper proposes an optimization-based stereo visual-inertial odometry (VIO) to locate indoor wheeled robots. Multiple Manhattan worlds assumption is adopted to model the interior environment. Instead of treating these worlds as isolated ones, we fuse the latest Manhattan world with the previous ones if they are in the same direction, reducing the calculated errors on the orientation of the latest Manhattan world. Then the structural lines which encode the orientation information of these worlds are taken as additional landmarks to improve positioning accuracy and reduce accumulated drift of the system, especially when the system is in a challenging environment (i.e., scenes with continuous turning and low textures). Besides, the structural lines are parameterized by only two variables, which improves the computational efficiency and simplifies the initialization of lines. Experiments on public benchmark datasets and in real-world environments demonstrate that the proposed VIO system can accurately position the wheeled robot in a complex indoor environment.

A NUMERICAL APPROACH TO THE CONTACT OF NOMINALLY FLAT SURFACES

Machined surfaces can be described by heights and wavelengths of the surface asperities that show a statistical variation. Considering that a regular wavy surface with a sinusoidal profile is the crudest model for a rough surface, studying the contact of regular wavy surfaces is a good approximation for the contact of nominally flat surfaces. Such contact problems exhibit periodicity that can be simulated with the aid of computational techniques derived for contact mechanics in the frequency domain. The displacement calculation, which is a necessary step in the resolution of the contact problem, is mathematically a convolution product that can be calculated in the frequency domain with increased computational efficiency. The displacement induced by a unit surface load can be expressed in the frequency domain by the frequency response functions, which are counterparts of the space domain solutions to half-space fundamental problems such as the Boussinesq problem. The displacement induced by a periodic pressure distribution can be computed by executing the convolution product between the frequency response function and pressure on a single period. It should be noted that the convolution calculation in the spectral domain implies that the contributions of all neighbouring pressure periods are accounted for. The need to treat numerically only a single period results in remarkable computational efficiency, allowing for high density meshes that can capture the essential features of any textured real surface. The displacement calculation promotes the solution of the contact problem by an iterative approach. The advanced method is benchmarked against existing analytical solutions for the 3D contact of surfaces possessing two-dimensional waviness. This essentially deterministic model, supported by a direct numerical solution that can be obtained for samples of real rough surfaces, presents itself as a worthy alternative to the existing statistical models for rough contact interaction.

Label GM-PHD Filter Based on Threshold Separation Clustering

Gaussian mixture probability hypothesis density (GM-PHD) filtering based on random finite set (RFS) is an effective method to deal with multi-target tracking (MTT). However, the traditional GM-PHD filter cannot form a continuous track in the tracking process, and it is easy to produce a large number of redundant invalid likelihood functions in a dense clutter environment, which reduces the computational efficiency and affects the update result of target probability hypothesis density, resulting in excessive tracking error. Therefore, based on the GM-PHD filter framework, the target state space is extended to a higher dimension. By adding a label set, each Gaussian component is assigned a label, and the label is merged in the pruning and merging step to increase the merging threshold to reduce the Gaussian component generated by dense clutter update, which reduces the computation in the next prediction and update. After pruning and merging, the Gaussian components are further clustered and optimized by threshold separation clustering, thus as to improve the tracking performance of the filter and finally realizing the accurate formation of multi-target tracks in a dense clutter environment. Simulation results show that the proposed algorithm can form a continuous and reliable track in dense clutter environment and has good tracking performance and computational efficiency.

Sensitivity of A Frequency-Based Alignment-Free Approach For Phylogeny Rreconstruction.

Background: Phylogenies enrich our understanding of how genes, genomes, and species evolve. Traditionally, alignment-based methods are used to construct phylogenies from genetic sequence data; however, this process can be time-consuming when analyzing the large amounts of genomic data available today. Additionally, these analyses face challenges due to differences in genome structure, synteny, and the need to identify similarities in the face of repeated substitutions resulting in loss of phylogenetic information contained in the sequence. Alignment Free (AF) approaches using k-mers (short subsequences) can be an efficient alternative due to their indifference to positional rearrangements in a sequence. However, these approaches may be sensitive to k-mer length and the distance between samples.Results: In this paper, we analyzed the sensitivity of an AF approach based on k-mer frequencies to these challenges using cosine and Euclidean distance metrics for both assembled genomes and unassembled sequencing reads. Quantification of the sensitivity of this AF approach for phylogeny reconstruction to branch length and k-mer length provides a better understanding of the necessary parameter ranges for accurate phylogeny reconstruction. Our results show that a frequency-based AF approach can result in accurate phylogeny reconstruction when using whole genomes, but not stochastically sequenced reads, so long as longer k-mers are used. Conclusions: In this study, we have shown an AF approach for phylogeny reconstruction is robust in analyzing assembled genome data for a range of numbers of substitutions using longer k-mers. Using simulated reads randomly selected from the genome by the Illumina sequencer had a detrimental effect on phylogeny estimation. Additionally, filtering out infrequent k-mers improved the computational efficiency of the method while preserving the accuracy of the results thus suggesting the feasibility of using only a subset of data to improve computational efficiency in cases where large sets of genome-scale data are analyzed.