An application of the eXtended Element-Free Galerkin method to a boundary-value problem reduces to an asymmetric EFG-type Saddle-Point (EFG-SP) problem. How-ever, saddle-point problems are difficult to solve even with iterative methods. For the pur-pose of resolving the difficulties, four types of high-performance solvers were developed for asymmetric EFG-SP problems in the previous study. In the four solvers, after elimi-nating Lagrange multipliers from the problems, the resulting linear systems are solved with Krylov subspace methods. In the present study, the Lagrange-multiplier-elimination method is generalized. As a result, an infinite number of solvers can be derived in principle. Three categories of solvers are introduced and their performance is investigated numerically. Con-sequently, it is found that the resulting solvers are effective especially for large-scale asym-metric EFG-SP problems.

We propose a method to rapidly generate matrices for real-symmetric eigenproblems. The proposed method produces a reproducible matrix with explicit eigenpairs, where the distribution of the eigenvalues can be controlled by a user. All elements of the generated matrix are rigorous floating-point numbers and can be represented in simple expressions involving the exact eigenvalues. The exact eigenpairs of the generated matrix are known in advance; thus, the proposed method contributes to the validation of errors in approximate eigenpairs. Several constraints on the matrix generated by the proposed method, were produced theoretically.

The purpose of this paper is to design a path following control system based on the kinematics of the Lizard-Inspired Single-Actuated robot (LISA). LISA is a new type of robot that mimics the quadrupedal walking morphology of lizards with a four-bar linkage mechanism and can realize both propulsion and turning with 1 degree-of-freedom. To achieve this purpose, this paper takes 3 approaches: kinematics formulation, control system design, and numerical simulation. In the kinematics formulation, we formulate LISA’s turning angle, stride length, posture, propulsive direction, curvature, and position coordinate. In control system design, we design a control system that converges not only distance error but also the posture error and control input. Conditional equations that can achieve these 3 control targets are formulated using forward kinematics and reference path functions, and control inputs are obtained by solving the inverse problem. Numerical simulations verify the effectiveness of the designed control system using 3 types of trajectories: linear, circular, and a combination of circular and linear trajectories. As a result, it is confirmed that the designed control system is effective for all 3 types of trajectories and that not only the distance error but also the attitude and control input can be converged.

Japanese traditional dance is a performing art with a rich and extensive history. Passed down orally and through demonstration over generations, its intricate movements are often challenging to convey accurately, resulting in subtle changes over time. To address this, efforts are being made to digitize and preserve these culturally significant practices through digital archiving. Furthermore, the number of traditional Japanese dance practitioners is declining yearly, leading to challenges in succession and education. This study employs two motion capture systems to create a digital archive of traditional Japanese dance and to visualize the movement differences between beginners and skilled practitioners. The ultimate goal is to develop a learning support system that aids in the education and training of future practitioners, ensuring the art form’s continuity.

In this study, we compare the parallel performance of the Improved Variable-Reduction Method (iVRM) and a variable preconditioned Krylov subspace method (VP Krylov) on a single GPU.We treat Poisson equations discretized by the element-free Galerkin method as target problems. iVRM avoids QR decomposition and eliminates Lagrange multipliers, enabling efficient solving of saddle-point systems dominated by sparse matrix operations. Numerical experiments show that both methods converge quickly, with iVRM exhibiting higher CPU efficiency. However, GPU results suggest that larger-scale problems are needed to fully utilize GPU resources. These findings guide the parallel design of iVRM for large-scale saddle-point problems.

This study examines the accuracy of numerical computations, focusing specifically on built-in elemental and special functions. The IEEE Standard 754 guarantees the accuracy of basic arithmetic operations and square root calculations. However, it does not establish precision criteria for built-in elemental and special functions, such as trigonometric, exponential, logarithmic, and gamma functions. In this paper, we evaluate the accuracy of low-precision elemental and special functions supported by MATLAB, GNU Octave, two C language compilers, CUDA, and NumPy in Python, using an interval enclosure technique. We assessed the accuracy of these functions for all low-precision floating-point inputs represented in FP16, BF16, and FP32. Additionally, we conducted a partial investigation into the accuracy of functions involving FP64 in Java within the aforementioned environment.

Previous studies have indicated that differences in the compatibility of singers in choral singing can be observed in terms of the first and second vocal tract resonant frequencies. However, the distinctions in the spectrograms of vocal harmonics remain unclear. In this study, we utilize a modified version of the Stockwell transform (MST), which features adjustable four parameters for optimizing time-frequency resolution. We attempted to select suitable parameter combinations by observing the effects of varying each parameter of MST through 3D plots of the indicator values. This approach enables us to clearly visualize vocal harmonics and reveal the synchronization of vibrato within the range of the frequency of the first and second vocal tract resonance, highlighting instances of good compatibility between singers.

It was demonstrated that the vortex field of a hybrid mode can propagate in the magnetized plasma region where plane waves are unable to propagate due to the cut-off condition. In this study, the dependence of the penetrated power of injected as millimeter-wave vortices of the hybrid mode in magnetized plasma is analyzed using the Finite-Difference Time-Domain (FDTD) method. The effect of the radius size of the corrugated waveguide on the penetrated power is described, revealing its significant contribution to reducing the deviation of topological charge in the hybrid mode. Furthermore, it was found that the penetrated power of the vortex field in magnetized plasma strongly depends on the topological charge l and the deviation of topological charge.

Characteristics of plasma radiation structure in radiative collapse were visualized using a two-dimensional radiation measurement and AutoEncoder (AE) on the Large Helical Device of the National Institute for Fusion Science, Japan. The state without collapse was treated as normal, and the state in which collapse is evolving was treated as abnormal. Using the anomaly detection by the AE, the collapse could be detected ~ 0.36 s before the collapse as increase in abnormality. Moreover, the abnormal radiation structure could be visualized as the profile of the reconstruction error which appeared ~ 0.46 s before the collapse from inboard side of the torus plasma.

We develop a deep learning model based on an extended pix2pix framework to predict 3D spatial distributions of X-ray dose to realize real-time monitoring of medical staff's radiation exposure. Utilizing conditional generative adversarial networks, the model processes 3D voxel grids to estimate X-ray dose distributions rapidly, addressing the limitations of Monte Carlo simulations in real-time applications. Training employed simulated datasets generated via a Monte Carlo code: PHITS. The trained model achieves an approximately 150,000-fold speedup compared to the Monte Carlo simulations. While the model predicts distributions with characteristics similar to the true values, errors increase in regions shielded by the objects.

This paper presents a novel reinforcement learning approach to enhance image matching, whereby the corresponding point candidates are detected from the feature points extracted from each of the two images. In general, robust estimation methods such as random sample consensus (RANSAC) are used to select valid corresponding points from the candidates. Therefore, we addressed the limitations of RANSAC random selection in the image-matching process, evaluating various reinforcement learning strategies, including deterministic and probabilistic approaches, and different value update mechanisms. The findings indicated that a probabilistic approach with suitable value updates provides a more robust solution for space-based navigation systems.

The power transfer efficiency of contactless power transfer devices that use vinyl-insulated wires was predicted by combining electromagnetic field analysis using the finite element method and equivalent circuit analysis using the Runge-Kutta method. We examined whether the measured power transfer efficiency is the median of the values obtained from calculations that considered vinyl-insulated wire to be a single solid wire and litz wire (vinyl-insulated wire is considered to be in an intermediate state between these two types of wire). It was found that the measured power transfer efficiency was not the median of the two calculated values in some cases. One of the calculated values was always slightly smaller than the measured value. The proposed method can thus be used to predict the lower limit of power transfer efficiency.

The pellet injection experiment in the tandem mirror device GAMMA 10/PDX is simulated with a plasma fluid model based on the anisotropic ion pressure (AIP model). A neutral gas shielding (NGS) model is implemented to model the pellet ablation rate. It is suggested that the energy loss due to charge exchange reactions promotes the axial particle loss from the main confinement region, i.e., the central cell. It is also found that the axial particle loss and energy loss owing to charge exchange reactions are the main contributors to the reduction in the stored energy in the central cell. To investigate the reason for the promotion of particle loss due to the charge exchange reaction, we analyze the force balance on the ion fluid. It is found that the particle flux at the edge of the central cell is determined by the pressure-gradient force, electric field force (in the direction of loss), and the mirror force (in the direction of confinement). In addition, it is shown that the mirror force decreases its absolute value more than the pressure-gradient force due to the energy loss caused by the charge exchange reaction, which enhances the particle loss from the central cell.

We propose a novel linear interpolation-based method to enhance the quality of widely-used JPEG images, which suffer from compression artifacts. Our approach focuses on restoring quantized DCT coefficients that have been rounded to zero during the JPEG quantization step. By employing a simple linear interpolation technique in the frequency domain, our method offers a computationally efficient solution with computational cost growing linearly with the number of restored coefficients. Experimental results on standard test images demonstrate that our method can improve the PSNR by 0.6 to 1.0 dB.

The magnetic fields confining plasma in fusion reactors are analyzed using Poincaré plots, which show intersections of magnetic field lines on a poloidal crosssection. Traditional methods for designing vacuum vessels and related structures involve slicing the reactor vertically and analyzing these plots, but this approach is inefficient for understanding plasma shapes globally. Key challenges include the computational cost of magnetic field line tracing and the difficulty in constructing continuous surfaces for the divertor legs due to the limited number of magnetic field lines reaching this area. To address these issues, a new and automatic method has been proposed to increase the number of magnetic field lines constituting the divertor legs by predicting optimal starting points for tracing. The proposed method involves placing starting points on orthogonal lines through the magnetic axis, resulting in a better representation of divertor legs. This new algorithm enhances the efficiency of generating Poincaré plots that depict divertor leg regions more clearly than previous methods. Neural Networks predict voxel data representing the shape of magnetic field lines. Considering the Larmor radius, we calculate an envelope surface that encompasses the region where plasma exists and create 3D modeling data.

Base station antennas are an indispensable component for mobile communication. For low wind load, circular cylinder structure with small diameter is usually employed for base station antenna. However, omnidirectional pattern with horizontal polarization and wide operating bandwidth is difficult to excite by such a cylinder structure. This paper presents the simulation and design method of Halo antenna with parasitic element for ease of fabrication and bandwidth enhancement. The proposed structure is Halo antenna with inner parasitic element. For ease of fabrication, several dimensions are fixed for the ease of fabrication and maintaining the inner capacity. The proposed antenna in this paper achieves operating bandwidth from 876 MHz to 975 MHz (10.8 %) with |S₁₁| under –10 dB. In addition, the omnidirectional pattern with horizontal polarization are confirmed in operating frequency band.

The “up3bd” code, which is an efficient three-dimensional (3D) electrostatic particle-in-cell (PIC) simulation code for study of transport dynamics in fusion boundary layer plasmas or other plasmas in nature, has been tested on various high-performance computing (HPC) systems which consist of different types of processor. The results of benchmark tests indicate that the up3bd code works faster on processors in which cache memory is larger or memory bandwidth is broader. Also, the types of computations that each architecture excels at and struggles with has been revealed.

A high-frequency electromagnetic field analysis based on the finite element method is known for poor convergence of iterative methods. Furthermore, false convergence may occur in which a physically correct solution cannot be obtained even though the convergence criteria are met. In this paper, we show that false convergence occurs in a high-frequency electromagnetic field analysis based on the finite element method. We propose and compare two methods to avoid false convergence that are currently executable: lowering the convergence judgment value of the residual norm of the conjugate orthogonal conjugate gradient method and increasing the wavelength resolution.

This study investigates the application of inchworm robots for mobile additive manufacturing (AM). A method for controlling the center of gravity (COG) during locomotion and designing unit segments for efficient fabrication is proposed. The relationship between link lengths and fabrication efficiency was analyzed, and unit segments were designed to balance stability and production efficiency. Simulations were conducted to evaluate the stability of the robot’s locomotion, confirming that the proposed design allows the COG to remain within the supporting leg area. These findings contribute to understanding the feasibility of inchworm robots as platforms for mobile AM systems.

The violin is known to withstand string tension and efficiently transmit string vibrations throughout the instrument due to the carefully carved arch shape, but there is still limited discussion and validation of analytical methods regarding the effects of string downward force on the overall structure. In this study, Finite element models of Willmotte Stradivari and Plowden Guarneri violin were developed, and numerical simulations were performed to visualize string downward force effects, and qualitative trends were consistent with measured data in the location and direction of displacement. As a result, it was suggested that the structural design may be related to the deformation caused by the string downward force effect.

Until now, authors have considered the Balancing Domain Decomposition DIAGonal scaling (BDD-DIAG) preconditioner as a preconditioner of non-overlapping domain decomposition analysis of 3-dimensional magnetostatic problems taking the magnetic vector potential as an unknown function. One interesting fact is that the direct solver without pivoting can be used in many cases for solving the coarse problem in BDD-DIAG. In this paper, they show a sufficient condition for positive-definiteness of the Schur complement matrix of the original magnetostatic problem in which the coefficient matrix of the finite element approximation is a well-known singular one. Authors only considered a well-known perturbed problem previously. But, in this paper, they consider the original magnetostatic problem. Furthermore, with a numerical example, they notice that the BDD-DIAG preconditioner includes the DIAGonal scaling (diag) preconditioner as its special case.

The structural coloration of pearls is due to optical multilayer interference by nacre. This paper considers the structural coloration of dark color pearls by nacre consisting of a multilayer of conchiolin and aragonite crystals. The Maxwell’s equations in 4×4 matrix form which use a relative permittivity tensor are introduced and solved. Calculating the reflection spectrum and the chromatic coordinates in standard red-green-blue (sRGB) color space, the effects of orientation of the c-axis on the structural coloration are investigated and shown.

Using machine learning, a method was developed to estimate the distribution of radiation sources arranged on a two-dimensional plane with high accuracy from the measured γ-ray energy spectra. A new machine learning method was used to construct a Spectrum Renormalization Filter (SRF) to convert the experimental data to be verified into a spectrum close to the shape of a simulation. The method using SRF produced more accurate estimation results than a method improving the accuracy of the simulation data used for training the estimation of the radiation sources distribution.