Kyutacar : Kyushu Institute of Technology Academic Repository
Not a member yet
9746 research outputs found
Sort by
Non-inverting Buck-Boost (NIBB) Converter Analysis of GaN Half Bridges Operation for 24V Unity Conversion Ratio
Get PDFDC-DC converters are essential for providing a fixed voltage to power LEDs and automotive systems. This paper introduces the non-inverting buck-boost (NIBB) DC-DC converter, which uses GaN half bridges to achieve a 24V unity conversion ratio operation. NIBB converter has been developed and tested for high-frequency feedback control under simulated LED load conditions. The feedback control was employed using FPGA and a high-speed ADC. The converter performance is examined for the high-frequency feedback control at switching frequencies of 500kHz and 800kHz corresponding to changes in the load condition. In addition, the efficiency performance for two different PCB designs of NIBB power converter circuit boards are compared under different operating conditions.journal articl
Underwater Image Recovery using In-situ Environmental Information
Get PDF九州工業大学九州工業大学博士学位論文（要旨）学位記番号: 生工博甲第512号 学位授与年月日:令和7年6月30日thesi
Consideration of The Difference in Accuracy Between Synthetic and Non-Synthetic Speech in Diarization
Get PDFIn recent years, speech segments detection technique called speaker diarization is becoming increasingly important, mainly for meetings, news, telephone speech and so on. However, conventional speaker diarization methods using neural networks require a huge amount of training data. We have shown that speaker diarization is possible with realistic training data by dividing relatively short speech signals recorded individually for each target person and synthesizing them on a computer. In this paper, we investigate the robustness of the method to speech recorded by microphones from multiple sound sources, and also examine the difference in accuracy depending on the position of the sound sources.journal articl
Code Reuse in CubeSat Projects: Analyzing the Adoption of Open-Source Software From the BIRDS Program
Get PDFCode reuse is a fundamental principle in software engineering, streamlining development processes and reducing costs. In the context of CubeSat development, open-source software frameworks, such as the Open-Source BIRDS CubeSat (OSBC) software, have the potential to enhance collaboration and accelerate mission deployment. However, the ex- tent to which CubeSat teams adopt and adapt these existing software solutions remains underexplored.
This study examines six CubeSat missions derived from the BIRDS satellite bus, assessing their software development strategies and the degree of code reuse from OSBC. By analyzing project documentation, source code repositories, and developer interviews, we investigate key factors influencing software reuse, including performance considerations, organizational constraints, and developer motivations. Our findings will provide insights into how open-source software impacts small satellite development efficiency and sustainability, highlighting best practices and barriers to adoption.
This research contributes to the broader discussion on knowledge reuse in space systems engineering and offers recommendations for fostering more effective software-sharing practices in future CubeSat missions.conference poste
Condition monitoring of a DC-link capacitor in an inverter with a front-end diode rectifier under imbalanced three-phase supply voltage
Get PDFDC-link capacitors in inverters have a shorter lifetime than the other devices, and thus degrade reliability of the inverters. The inverters are usually fed by three-phase supply voltages; however, the three-phase voltages are frequently imbalanced due to the connection of single-phase power sources and loads, such as renewable energy, which places additional stress on the DC-link capacitors. This paper proposes a condition monitoring method of a DC-link capacitor without an additional current sensor in an inverter system under the imbalanced three-phase supply voltage. This inverter system employs a front-end six-pulse diode rectifier with a DC reactor. The method is based on an analysis of the rectifier output ripple voltage including the uncharacteristic harmonics that result from imbalanced supply voltage, which is valid in a practical imbalance ratio around 5%. Experimental results obtained from a 200-V 1.5-kW laboratory system confirmed that both the capacitance and ESR were monitored even though the supply voltage was imbalanced.journal articl
CFD Analysis of Flow near Shark Skin Model with Vibration for Improvement of Antifouling Properties of Ship Bottom: Analysis of Flow near Multiple Denticles that Compose Shark Skin
Get PDFresearches related to bio-inspired shark skin’s drag reduction and antifouling to improve fuel consumption of ships. But the bio-inspired shark skin has a demerit that antifouling effect of skins is lost when they are at anchor in a few weeks. In this study, to add the antifouling effect on the bio-inspired shark skins at anchor, the vibration of the skin is tried to be used. As for the targeted model, real shark skin model is created by using µCT figures of mako shark skin. Flow field around the shark skin is analyzed by CFD when real shark skin or flat skin vibrate. The fluid is water and it is treated as laminar flow. To simulate flow field by sinusoidal vibration of shark skin, opposite side to skin is treated as inlet and shark skin side except for shark skin is treated as outlet. As for the boundary condition for these parts, velocity and pressure are given as sinusoidal wave at the inlet, and pressure is 0 at the outlet. And boundary condition at shark skin and other sides are treated as fixed wall. In addition, wall shear stress on one denticle and two denticles is analyzed. As a result, it is shown that flow field near shark skin is complicated and maximum wall shear stress on the skin increases as frequency of vibration increases. And there is linear relationship between flow velocity and wall shear stress. In the case of vibration of two denticles of shark skin, wall shear stress on the skin is influenced by distance and angle between two denticles. It is concluded that (1) The possibility of adding antifouling effect on shark skin increases with vibration, (2) Wall shear stress on shark skin will be increased by optimizing distance and angle between two denticles in the case of vibration.journal articl
The Role of Life Cycle Assessments in Digital Product Passport Implementation for Building a Plastic Circular Economy
Get PDFThe digital product passport (DPP) system is expected to support circularity, transparency, and traceability in the circular economy. However, environmental impacts posted on the DPP necessitate a mandatory life cycle assessment (LCA). This paper discusses the construction of a plastic circular economy system from the perspectives of DPP and LCA. The status of the DPP in the European Union (EU) is reviewed, and a project on the technical development of the DPP in Ja-pan using plastics as a case study is introduced. We also explain the role of LCAs in the circular economy and discuss the necessity, challenges, and prospects of LCA, considering the quality of plastic waste, in order to withstand resource circulation. Our study reveals the need for LCA in the determination of the environmental impacts of various combinations of methodologies and technologies applied during each process of the circular economy of plastics. This allows manu-facturers and consumers to select plastic products with low environmental impacts using DPP information.journal articl
Nurse Activity Recognition in Gastrostomy Tube Feeding Using Video-Based Pose with Large Language Model-Guided Features
Get PDFIn this paper, we improve nursing activity recognition in gastrostomy tube feeding (GTF) with temporal variations and sequential errors by integrating activity context to Large Language Model (LLM) for guided feature selection and post-processing. GTF is a delicate nursing procedure that allows direct stomach access in children for supplemental feeding or medication, but it is underrepresented in datasets, posing challenges for accurate detection. Manual feature engineering may overlook subtle but important motion cues, particularly in opening and closing the gastrostomy cover, where changes are minimal and localized to the hands. Additionally, sequence inconsistencies and missed activities limit the effectiveness of pose estimation methods in healthcare. Leveraging the contextual adaptability of LLMs, we generate new features suggested by the language model, combining them with hand-crafted features to optimize the model. For post-processing, a sliding window smoothing method based on majority voting is applied. To mitigate duration-based discrepancies, a priority handling is incorporated for short-duration activities to pre- serve their recognition accuracy while addressing repeated labels caused by long-duration actions. Particularly, we applied activity recognition to our unique GTF dataset collected from recorded video of two nurses, two students, and two staff members for three days with 17 labeled activities. Keypoints are extracted using YOLO11. Compared to the baseline, the application of LLM to GTF nurse activity recognition with pose estimation improved the Random Forest performance of F1-score from 54% to 57%. Additionally, incorporating the sliding window smoothing approach based on majority voting with short-term action priority, resulted in a 3% further increase.journal articl
Summary of the Virtual Data Generation for Complex Industrial Activity Recognition
Get PDFThis paper presents a summary of the virtual data generation for complex industrial activity recognition Challenge, which focused on exploring virtual data generation techniques to improve the performance of human activity recognition (HAR) in complex industrial environments. The challenge utilized the OpenPack dataset, a large-scale multimodal collection of sensor data captured during real-world packaging operations. Participants were tasked with generating synthetic accelerometer data to augment a baseline HAR model. Four teams from different countries proposed diverse approaches, including interpolation, classical augmentations, variational autoencoders, and GAN-based methods. Their submissions were evaluated using micro F1 score across multiple random seeds to test robustness. The results reveal that while deep generative models offer strong potential, simpler signal-based techniques also perform competitively when wellaligned with the data structure. Additionally, incorporating finer-grained action labels within each operation can help guide more realistic virtual data generation, leading to improved HAR model performance by better capturing intra-operation dynamics. Based on these findings, we discuss key insights and suggest future directions for designing robust, semantically consistent, and computationally efficient virtual data generation pipelines for industrial HAR applications.journal articl
