Kyutacar : Kyushu Institute of Technology Academic Repository
Not a member yet
9746 research outputs found
Sort by
Lam'e equation on a circle and application to singular limit eigenvalue problems
Get PDFLet n ∈ {1,2,．．．}, l ∈ {1,2¥} and 0<k<1. We are concerned with the eigenvalues of the Lamé equation of the form
-Φ''+l(l+1){e3+k2sn2(x,k)}Φ=λΦ for x ∈ R/2nK(k)Z,
where e3:=-(1+k2)/3 is a constant, sn(x,k) denotes Jacobi's elliptic function and K(k) denotes the complete elliptic integral of the first kind. Using integral representations of two independent solutions, we obtain exact expressions of all eigenvalues and establish asymptotic formulas for all eigenvalues as k → 1.
It is known that the Lamé equation appears as a linearized eigenvalue problem of important semilinear elliptic equations including the Allen-Cahn equation, a scalar field equation and the sine-Poisson equation.
We also establish asymptotic formulas for the eigenvalues of the linearization of various boundary value problems of semilinear elliptic equations.journal articl
Environment Recognition in Spherical Video Images Using Multi-Attention DeepLab
Get PDFDue to the aging population in Japan, the demand for electric wheelchairs is increasing accordingly. To reduce the cost of an autonomous recognition of environment, the spherical camera is widely used to replace the expensive LiDAR, which can also provide a 360-degree view with a single device. Automatic environment recognition has become an auxiliary method to reduce traffic accidents and improve the driving experience of electric wheelchairs. However, environment recognition remains a major challenge due to the severe distortion of images from spherical cameras. In this paper, we propose a convolutional neural network-based method for semantic segmentation in the field of image recognition. Multiattention pyramid upsampling is introduced to improve accuracy by dealing with image distortion. Our method was applied to the SYNTHIA dataset and to a dataset created by cropping images from videos recorded at our campus. The mean intersection over union (mIoU) was 72.2%, and the processing speed was 28.2 frames per second in the combined dataset.journal articl
Super-Resolution Reconstruction of Satellite Images Using Refined SwinIR
Get PDFWith the explosion in the number of low-cost satellites, satellite imagery has become widely used for a variety of applications, including agriculture, landscape, and environmental awareness. Therefore, one of the immediate challenges is to increase the resolution of images to extract useful information. Accordingly, higher resolution satellite images are expected to improve recognition accuracy. On the other hand, there are several possible limitations to the use of satellite imagery. Therefore, there is a need to develop a method that can increase the resolution of images from as little satellite image data as possible. In this study, we developed a deep learning model with a small number of parameters. We developed Refined SwinIR, an improved version of the conventional method that reduces the number of parameters while maintaining the quality of the results. Satellite images from Google Earth Pro were used for training and testing. Compared to methods based on classical convolutional neural network modules, the number of parameters used in our model was reduced by 44.45%, and we were also able to improve the peak signal-to-noise ratio and structural similarity values.journal articl
Limonene Selectively Modulates Visual Attention Through P300 Suppression: A Comparative Event-Related Potential Study With Lemon Essential Oil
Get PDFIntroduction
Odor stimuli can influence cognitive processes, including selective attention. However, whether these effects are driven by the chemical properties of odor compounds or by their semantic or emotional associations remains unclear. This study aimed to clarify the mechanism by which olfactory inputs influence visual attention by isolating the effects of specific odor compounds from those of affective or semantic factors.
Methods
To investigate this question, we assessed the effects of limonene, a key compound found in citrus odors, and lemon essential oil on visual attention, as indexed by the P300 component of event-related potentials, which reflect the allocation of selective attention. Participants completed a visual oddball task under three odor conditions: no odor, limonene, and lemon essential oil.
Results
Limonene presentation significantly reduced P300 peak amplitudes compared with the no-odor condition, whereas lemon essential oil showed not significantly reduce. Subjective ratings of pleasantness, congruency, and arousal did not differ significantly between the two odors.
Conclusions
Our results suggest that limonene modulates selective attention primarily through its chemical properties rather than through mood or semantic influences. This study provides new evidence for chemically specific olfactory–visual interactions and underscores the importance of distinguishing chemical effects from psychological mechanisms in cross-modal cognitive modulation.journal articl
FPGA-based Sensorless Control Strategy for Permanent Magnet Synchronous Motor using High-Frequency Injection with Loaded Start-up Capability
Get PDF九州工業大学博士（工学）1 Introduction| 2 The control principle of a permanent magnet synchronous motor| 3 Sensorless control strategy for permanent magnet synchronous motor based on saliency tracking| 4 EMA-based Position Error Signal Extraction in the HF injection technique| 5 A suitable inverter-fed drive designed for sensorless permanent magnet synchronous motor drives| 6 FPGA-based sensorless permanent magnet synchronous motor drive systems| 7
ConclusionsThis dissertation proposes an FPGA-based sensorless control strategy for Permanent Magnet Synchronous Motors (PMSMs) utilizing saliency-based tracking in the High-Frequency (HF) injection technique to achieve accurate rotor position estimation, particularly in the low-speed region. A significant challenge in PMSM control lies in ensuring reliable, high-performance operation at low speeds, where traditional encoders or mechanical sensors face limitations. From standstill to low speeds, these sensors suffer from reduced resolution, weaker signal quality, insufficient sampling rates, and increased mechanical play and friction, all of which contribute to inaccuracies in speed and position detection. In addition, start-up conditions from standstill under load remain a critical focus in industrial motor control research, as many industrial applications, including electric vehicles, milling systems, and elevators, demand reliable operation in these scenarios. To address these challenges, this study presents fast processing by FPGA for a real-time sensorless control approach using the HF injection technique with start-up capability under loaded, eliminating the need for a speed sensor. Rotor position is estimated through modulation signals in the quadrature current, induced by a HF signal injected into the d-axis, which generates an estimated coordinate system to determine angular speed differences between the real and estimated systems. Digital filters and analytical calculation within modulation process and extraction scheme related to rotor position information, as well as loop-control tracking by extended Proportional-Integral (PI) control, combined with differential equations, are employed to accurately track and estimate rotor position and speed. Moreover, circuit design also plays a vital role in reducing noise, a common issue during sensorless low-speed operation. The integrated design of the motor control system, including the gate driver, power inverter, and control processes, should be considered to effectively mitigate noise and eliminate undesirable behavior of power switching components. Ringing suppression circuit is employed to improve noise reduction, providing precise current signals for modulation techniques, enabling accurate rotor position and speed estimation. Finally, experimental validation demonstrates the success of the proposed method, achieving reliable start-up operation under load, even at speed below 1 min-1.九州工業大学博士学位論文 学位記番号:生工博甲第514号 学位授与年月日:令和7年9月25日令和7年度doctoral thesi
Vision-Based Human Intention Recognition and Robotic Assistance for Smart Work Cells
Get PDF九州工業大学博士（工学）1 Introduction| 2 Related Work| 3 Proposed Method| 4 Proposed System| 5 Experiments and Results| 6 Discussion and ConclusionHuman-robot collaboration (HRC) plays a vital role in enhancing productivity and inclusivity in modern work environments. However, existing systems often fail to address the specific needs of individuals with physical disabilities. This dissertation proposes an intelligent, vision-based robotic assistance system tailored for inclusive workcells, specifically supporting users with only one functional arm. The system integrates a depth camera, a gimbal-based motion tracking mechanism, and MediaPipe-based human pose and hand tracking, combined with lightweight machine learning models for real-time human intention recognition.
The robotic platform centers around a UR3e collaborative arm equipped with modular, 3D-printed end-effectors designed for flexible object handling and soldering tasks. The system supports both manual and autonomous operation modes, enabling seamless gesture-to-action mapping without requiring wearable devices. Real-time gesture classification is achieved with high accuracy using ensemble learning techniques, demonstrating over 90% recognition rates in dynamic settings.
Comprehensive experiments validate the system’s robustness across diverse user conditions, including gesture variability, spatial movement, and real-time task execution. A gimbal-stabilized vision pipeline ensures consistent tracking, and safety features are integrated to prevent unintended robot actions. While initial tests were conducted with healthy users, the system is designed for older adults and persons with disabilities, and future studies will expand its evaluation in those populations.
This work bridges theoretical advancements in vision-based intention recognition and practical deployment of assistive robotics. It contributes a scalable and adaptable framework for inclusive automation, addressing critical challenges in aging societies and paving the way for future human-centered manufacturing.九州工業大学博士学位論文 学位記番号:生工博甲第518号 学位授与年月日:令和7年9月25日令和7年度doctoral thesi
持続可能な住民参加型道路維持管理制度の仕組みと地域コミュニティ活性化に関する研究
Get PDF九州工業大学博士（工学）1 序論| 2 既往研究と本研究の位置づけ| 3 北九州市道路サポーター制度の概要| 4 住民参加型道路維持管理活動に有効な行政支援| 5 制度運営における中間支援組織の機能とそれを支える仕組み| 6 制度を通じた多様な地域コミュニティ強化の可能性| 7 総括九州工業大学博士学位論文 学位記番号:工博甲第600号 学位授与年月日:令和7年3月25日令和7年度doctoral thesi
Elucidation of Impulse Voltage Degradation Mechanism in Cast Resin Transformers
Get PDF九州工業大学博士（工学）1 Introduction| 2 Measurements of AC Partial Discharge of an Aged Cast Resin
Transformer after Impulse Voltage Applications| 3 Partial Discharge and Degradation Characteristics of Cast Resin Transformers Due to Bipolar Impulse Application| 4 Degradation Causes of Twisted Pairs Due to Bipolar Impulse Application| 5 Degradation Mechanism caused by Impulse Partial Discharges in the Void within the Polyethylene Terephthalate Films| 6 Conclusions and Future WorkCast resin transformers (CRTrs), characterized by windings insulated with epoxy resin, are widely used in modern power distribution systems due to their superior performance. However, an increasing number of partial discharge (PD)-related electrical failures have been reported involving CRTrs operating beyond 20 years, underscoring the need for an in-depth investigation into their long-term degradation mechanisms. In particular, the aging mechanisms under high impulse voltages, such as those caused by lightning or switching overvoltages, remain insufficiently understood, and the lack of reliable diagnostic tools complicates the prediction of remaining service life and maintenance planning.
This dissertation addresses these challenges through experimental studies aimed at elucidating the insulation degradation mechanisms induced by impulses in CRTrs.
Chapter 2 investigated PD behavior in an aged 22 kV CRTr subjected to multiple impulse voltages of varying magnitudes, polarities, and cycles. PD inception voltage, PD maximum charge, and phase-resolved partial discharge patterns were analyzed. Results show that insufficient impulse magnitudes did not significantly accelerate insulation degradation, prompting further investigation.
Chapter 3 studied systematically investigates the PD behavior and aging characteristics of CRTr windings under multiple bipolar impulse. A higher-magnitude impulse voltage application and impulse PD detection systems were developed to study impulse PD generation, and development, using a high-frequency current transformer. Impulse PD magnitude was precisely calculated by the current integration method. Results indicate that higher-magnitude impulse voltages cause impulse PD activity in the windings. A three-stage mechanism, space charge accumulation, residual charge deposition, and breakdown, is proposed to describe the degradation process. Impulse PD-induced space charge accumulation and residual charge deposition significantly accelerate insulation aging and cause irreversible degradation.
Chapter 4 extended the investigation to twisted-pair samples coated with epoxy resin, simulating turn-to-turn insulation. AC PDIV and capacitance of the twisted pair samples were measured before and after bipolar impulse applications. Experiments showed that changes in AC PDIV and capacitance indicate aging trends. Microscopic analysis revealed the existence of air voids, a key indicator of insulation degradation, further supporting insulation degradation mechanisms induced by an air cavity.
In Chapter 5, a polyethylene terephthalate film model with an artificial void was designed. The role of residual charges in the subsequent discharge was elucidated by using surface charge mapping method. Observations of surface morphology changes, and AC PD characteristics were analyzed to provide insights into the mechanisms of insulation degradation. Experiments demonstrated that residual charges influence PD characteristic, with bipolar impulses accelerating aging more effectively than unipolar impulses. Surface charge mapping quantitatively assessed residual charge accumulation and their relationship with impulse PDs, revealing its influence mechanism on the development of and the subsequent PD behaviors. A decrease in the steady PDIV was observed due to impulse PD exposures. This research concludes that impulse PD induces irreversible insulation degradation through polymer chain scission and surface erosion, decreasing PDIV and increasing vulnerability to breakdown.
The main contributions and innovations of this work are: 1. Development of a PD detection method under high-magnitude impulses. 2. Proposal of a three-stage degradation mechanism driven by space charge and residual charge dynamics. 3. Introduction of surface charge mapping to quantify residual charges within voids and their effects on the subsequent PD behavior. 4. Elucidation of degradation mechanism caused by impulse PDs through molecular chain scission and surface erosion, leading to irreversible insulation degradation.
The outcomes contribute to paving the way for the accelerated aging tests in the windings of the CRTr by bipolar impulse voltages, emphasizing the dangers of bipolar switching overvoltages or lightning overvoltages on the insulation status of the transformers in the power systems, clarifying the hazards of the presence of air cavities on solid insulation transformers, proposing residual charge and space charge as an indicator for insulation monitoring, guiding the development of advanced diagnostic methods based on the measurement of the steady PDIV to predict the operational lifetime of CRTr and enhance power system reliability.九州工業大学博士学位論文 学位記番号:工博甲第601号 学位授与年月日:令和7年3月25日令和7年度doctoral thesi
車載用コモンモードチョークコイルによる妨害波抑制効果の簡易評価法に関する研究
Get PDF九州工業大学博士（工学）1 序論| 2 従来研究の調査と本研究の課題| 3 車載向けコモンモードチョークコイルのミクスドモードS パラメータを活用した放射エミッションへの影響調査| 4 等価コモンモード電圧の検討と放射エミッションの影響の簡易推定検討| 5 等価コモンモード電圧測定法の検討| 6 結論九州工業大学博士学位論文 学位記番号:工博甲第603号 学位授与年月日:令和7年3月25日令和7年度doctoral thesi
Atomic and electronic structures of point defects related to vacancy-mediated mechanism of Mg diffusion in GaN
No full textWe report first-principles total-energy calculations based on the density functional theory (DFT) with hybrid approximation to the exchange-correlation functional, which provide a firm theoretical framework to understanding the fundamentals of certain point defects, i.e., the substitutional Mg at the Ga site (MgGa), the Ga vacancy (VGa), and the pair of the substitutional Mg and the Ga vacancy (MgGa-VGa) in GaN. These defects play an essential role in the Mg migration in GaN. We find a competing feature between the spin polarization peculiar to the localized N orbitals and the crystal-field splitting common to tetrahedrally bonded semiconductors. As a peculiar exhibition of such competition, we find (i) structural multi-stability of MgGa, (ii) charge-state-dependent diversity in spin polarization in VGa, and (iii) binding of MgGa and VGa due mainly to the electrostatic attraction caused by electron transfer among N orbitals around the pair.journal articl