Kyutacar : Kyushu Institute of Technology Academic Repository
Not a member yet
9746 research outputs found
Sort by
Data Analysis and Visualization Method for Equipment Anomaly Detection by Using Acoustic Data
Get PDF一般的な設備機器の手動点検においては，労働力や時間，人件費といった多くの課題がある．これに加え，点検と故障のタイミングのずれは適切なメンテナンスを困難にすることから，設備の安定的な運用に悪影響を及ぼす可能性がある．本研究では，このような課題を解決するために，変電所の油圧ポンプを対象として機器の駆動音を長期的に録音し，そのデータを可視化するシステムを開発する．また，同時に油圧ポンプの駆動時間と気象データの相関を分析し，その結果から安全な駆動時間範囲や故障の予兆を判断できるか検証する．これにより，設備機器の安定的な運用に寄与し，手動点検に依存せずに適切なメンテナンスを実現する手段を提供する．本システムの有効性を評価するため，11 台の油圧ポンプを対象にデータ収集，分析，そして可視化までの一連の動作を検証し，相関分析，線形回帰分析，残差分析の結果から機器の駆動安全範囲を推定した．Manual inspections of general equipment present many challenges in terms of manpower, time, and labor costs. In addition, the timing gap between inspections and breakdowns makes proper maintenance difficult, which can have a negative impact on the stable operation of equipment. In order to solve these problems, this study develops a system to record the driving noise of hydraulic pumps in substations over a long period of time and visualize the data. At the same time, we will analyze the correlation between hydraulic pump operation time and weather data, and verify whether we can determine safe operation time ranges and predictive signs of failure from the results. This will contribute to the stable operation of equipment and provide a means to achieve appropriate maintenance without relying on manual inspections. To evaluate the effectiveness of the system, a series of operations from data collection, analysis, and visualization were verified for 11 hydraulic pumps, and the safe operating range of the equipment was estimated from the results of correlation analysis, linear regression analysis, and residual analysis.電子情報通信学会 インターネットアーキテクチャ研究会（IA）, 2024年3月12日-14日, 宮古島市未来創造センターjournal articl
Facile and scalable pulsed electrodeposition of bi-active Bi-Sb alloy for high performance sodium ion batteries
No full textAlthough sodium ion batteries (SIB) have shown great potential for large-scale energy storage systems, the development of high-performance anode materials for SIB is crucial for their progress. However, the cycling performance of SIB is currently limited by severe volume changes during the sodiation/desodiation process. To overcome this problem, a bi-active metal alloying strategy has been proposed. In this study, we utilized pulsed electrodeposition to prepare bismuth-antimony alloy anode materials. The resulting Bi44.9Sb55.1 alloy anode exhibited exceptional specific capacity, rate performance, and stability. After the second cycle, the discharge/charge capacities were 505.3 mAh·g−1/496.7 mAh·g−1 at 1 A·g−1, with a coulombic efficiency of over 98 %. Even after 50 cycles, the capacities remained at 494.7 mAh·g−1/481.9 mAh·g−1 with a capacity retention rate of 97.9 %. These excellent properties were attributed to the stable structure and suitable voids for electron transport and ion diffusion in the Bi44.9Sb55.1 alloy, which effectively mitigated volume expansion during cycling. The reaction kinetics and mechanism were studied using cyclic voltammetry, in situ X-ray diffraction, and nuclear magnetic resonance techniques. Our work provides valuable insights for the development of bi-active metal alloy anodes for SIB.journal articl
An Integration of Contact Force Models with Multibody Dynamics Analyses for Human Joint Mechanisms and Effects of Viscoelastic Ground Contact
Get PDFIn human movement and rehabilitation analyses, human joint dynamics is a key to consider the incorporation of spring-damper components, flexible bodies and contact forces analytically. In the present study, an analytical method for human gaits were introduced to integrate those essential elements, and viscoelastic properties of musculoskeletal system were modeled with the absolute nodal coordination formula (ANCF) method representing flexible body motions. A contact force model simulates interactions between different body segments and the environment. The proposed system is applied to a slider crank mechanism, demonstrating its capabilities in human joint motion analysis using integrated dynamic model within the framework of multi body dynamics (MBD), which realizes dynamic/inverse dynamics for human biomechanics. © The 2024 International Conference on Artificial Life and Robotics.The 2024 International Conference on Artificial Life and Robotics (ICAROB 2024), February 22-25, 2024, on line, Oita, Japanconference pape
Stand-Up Indulgent Gathering on Lines for Myopic Luminous Robots
Get PDFAbstract We consider a strong variant of the crash fault-tolerant gathering problem called stand-up indulgent gathering (SUIG), by robots endowed with limited visibility sensors and lights on line-shaped networks. In this problem, a group of mobile robots must eventually gather at a single location, not known beforehand, regardless of the occurrence of crashes. Differently from previous work that considered unlimited visibility, we assume that robots can observe nodes only within a certain fixed distance (that is, they are myopic), and emit a visible color from a fixed set (that is, they are luminous), without multiplicity detection. We consider algorithms depending on two parameters related to the initial configuration: Minit, which denotes the number of nodes between two border nodes, and Oinit, which denotes the number of nodes hosting robots. Then, a border node is a node hosting one or more robots that cannot see other robots on at least one side. Our main contribution is to prove that, if Minit or Oinit is odd, SUIG can be solved in the fully synchronous model.Proceedings of the 38th International Conference on Advanced Information Networking and Applications (AINA-2024), 17-19 April,2024, Kitakyushu, Japanjournal articl
A Cognitive and Constructivist Approach to Evidentiality — From the Perspective of Continuity Between Grammar and Lexicon —
Get PDF本論文は「九州地区国立大学教育系・文系研究論文集」Vol.10,No.2に査読を経て受理された。journal articl
SSdetector: Secure and Manageable Host-based IDS with SGX and SMM
Get PDFHost-based intrusion detection systems (HIDS) are used to monitor the internals of target systems. It is essentially difficult to execute HIDS securely inside target systems. For example, it is not guaranteed that HIDS can obtain correct information from compromised systems. If HIDS is tampered with by intruders, it would be easily disabled. So far, various techniques have been proposed to securely execute HIDS using the security features of processors, e.g., System Management Mode (SMM) and SGX in Intel processors. However, strongly secure HIDS sacrifices its manageability, whereas manageable HIDS is less secure. In practice, it is important to achieve not only the security but also the manageability of HIDS. This paper proposes SSdetector for achieving both security and manageability by combining SGX and SMM. SSdetector securely runs HIDS inside an SGX enclave, which is a protected region inside an SGX application. Since HIDS is developed as an SGX application, the management of HIDS is easier. To securely obtain system information in memory, in-enclave HIDS invokes the SMM monitor running in an isolated execution environment created by BIOS. SSdetector protects information passed between in-enclave HIDS and the SMM monitor by encryption and integrity checking. We have implemented SSdetector in UEFI BIOS and examined the performance of HIDS collecting system information necessary for the proc filesystem.2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom), 01-03 November, 2023, Exeter, United Kingdomjournal articl
Time-domain Analog VLSI design of Neural Network Circuits Assuming a Use of Analog Memory Devices with Non-ideal Properties
Get PDF5th International Symposium on Neuromorphic AI Hardware, March 1-2, 2024, Kyushu Institute of Technologyconference pape
Optoelectronic deep neural network hardware using principle of self-referential holography
Get PDFSelf-referential holography (SRH) is a purely one-beam holographic data recording geometry based on a volume hologram recorded by self-interference of a spatially phase-modulated light. We have proposed to apply the principle of SRH for a holographic data storage without use of a reference beam, which is named self-referential holographic data storage (SR-HDS). Also, recently, we have found that SRH can be used as a deep neural network hardware by combining it with electronic calculations. In this paper, we mainly introduce the self-referential holographic deep neural network (SR-HDNN) and show proof-of-principle simulation results of 4-class image recognition task.journal articl
Use and implementation of deep learning in self-referential holographic data storage
Get PDF自己参照型ホログラフィックメモリ（SR-HDS）は，ホログラフィックメモリ（HDS）の実装形態の一つであり，参照光を用いないことを特長とする．参照光を用いないため光学系が安定ではあるが，設計の自由度が低いという側面もある．このことは，SR-HDSの記録再生品質を向上させるという目的において，光学設計によるアプローチには限界があることを示唆している．そこで本研究では，近年注目されている深層学習を用いるアプローチに着目する．具体的には，記録再生光の位相分布の設計，再生データページのデノイズおよびブロック符号の復号などに深層学習を適用することができると期待される．さらに我々は，深層学習に必要な一部の演算をSR-HDSに記録されているホログラムと少しの電子処理に担わせる実装方法，つまり，深層学習の機能を内在するSR-HDSの確立を目指している．本稿では，SR-HDSとSR-HDSの光学系を用いた深層ニューラルネットワーク（DNN）ハードウェアの実装方法である自己参照型ホログラフィック深層ニューラルネットワーク（SR-HDNN）を紹介する．最後にそれらの融合に対する期待と展望を述べる．Self-referential holographic data storage (SR-HDS) is one of the implementations of holographic data storage (HDS), and does not require a reference light which is required in conventional HDS systems including two-beam and collinear geometries. By this feature, the optical system of SR-HDS is enabled to be stable and compact, however the degree of freedom of optical design is limited. This means that the significant improvement of the recording qualities by optical design approach is not expected. Therefore, we have focused on approaches using deep learning which have been attracting attention in recent years. For example, we expect to use deep learning-based approaches in design of the phase pattern used in recording and reading processes, denoising of the reconstructed datapages, and decoding block codes. In this paper, we introduce SR-HDS and how to implement deep neural network (DNN) hardware by the same optical system as SR-HDS, which is named self-referential holographic deep neural network (SR-HDNN). Finally, we discuss how to establish SR-HDS internalizing deep learning capabilities.HODIC2024年第1回研究会，2024年3月25日，千葉県千葉市（ハイブリッド）journal articl
Computational Analysis of Receptive Fields of Retinal Ganglion Cell During Fixational Eye Movements
Get PDF5th International Symposium on Neuromorphic AI Hardware, March 1-2, 2024, Kyushu Institute of Technologyconference pape