14532 research outputs found
Sort by
A Finite State Automaton for the Tsetlin Machine Using High-Capacity Resettable D Flip-Flop Cells
No full textSingle Flux Quantum (SFQ) circuits provide ultrafast operation and low power consumption, making them attractive candidates for machine learning accelerators. The Tsetlin Machine (TM) is a low-complexity and highly interpretable architecture, organized as a voting system of propositional clauses. In the TM, an ensemble of finite-state automata (FSAs), called Tsetlin automata, independently decides whether to include or exclude specific literals. The chosen literals are assembled into propositional clauses that collectively form the voting system, while each FSA updates its state based on rewards or penalties determined by the prediction outcome. However, existing FSA designs for TM based on SFQ circuits still suffer from either a narrow margin or excessive circuit area. To achieve compact and stable state transitions using SFQ circuits, we designed High-Capacity Resettable D Flip-Flop (HC-RDFF) cells by cascading storage loops. The proposed FSA was implemented using two HC-RDFFs to store the include and exclude states, respectively. The FSA with 16 states requires only 395 Josephson junctions with a static power of approximately 128 W, exhibiting a bias margin of −21% to +25% at low speed and a maximum operating frequency of 20 GHz in simulation. We fabricated the HC-RDFF and the FSA using a 10 kA/cm² Nb four-layer superconducting process and verified their correct operation through low-speed measurements. The number of FSA states can be increased simply by extending the cascaded storage loops in the HC-RDFFs. We expect the proposed FSA to be suitable for constructing stable TM systems based on SFQ circuits with reduced hardware costs
Gas phase and surface processes of SiCxNyOz film formation by parallel plate plasma-enhanced chemical vapor deposition without heat assistance
No full textThe three steps to quantitatively evaluate the gas phase and the surface processes were developed for a SiCxNyOz film formation using the parallel plate plasma-enhanced chemical vapor deposition without heat assistance. (A) The measured film thickness was expressed by the thickness equation using the fourth-degree polynomial forms of the mathematical product of the electric current, I, and the partial pressures, P, of the monomethylsilane, nitrogen and argon gases. (B) The data set consisting of the film thickness and the film formation conditions in the typical thickness range was prepared using the thickness equation. (C) The data set was analyzed assuming the gas phase and the surface processes, such as (i) the multi-order reactions of the nitrogen and the monomethylsilane gases in the gas phase to produce the intermediate compound, (ii) the chemisorption of the intermediate compound at the surface, (iii) the decomposition of the chemisorbed intermediate compound by the argon gas to produce the SiCxNyOz film and (iv) the etching of the SiCxNyOz film surface by the argon gas. The film thickness was sufficiently reproduced by the rate equation in this study. The surface coverage by the intermediate compound was evaluated to be 0.02-0.1 %. The influence of the argon etching was significant, because only 8-18% of the total formed thickness remained as the film
Dynamic evaluation method for assessing lateral stability of ballasted railway tracks with defects
No full textLateral stability is essential for the safe and efficient operation of ballasted railway tracks, ensuring proper alignment and reducing derailment risks. Conventional evaluation methods, such as single-sleeper pullout tests, are time-consuming and lack scalability. In this study, we propose a novel dynamic loading evaluation method for the rapid assessment of lateral stability in ballasted tracks with various defects. Using 1/9-scale track models, we examined the effects of reduced ballast density, localized ballast loss, and sleeper settlement on the lateral resistance through both static and dynamic tests. In the static tests, lateral resistance was measured by pulling a central sleeper under controlled conditions. The results demonstrated a clear correlation between the ballast condition and resistance. For example, the tracks with a ballast density of 1.4 g/cm³ showed approximately 50% lower resistance than those with 1.6 g/cm³. Similarly, significant reductions could be observed or up to 100% ballast loss and five unsupported sleepers. The dynamic tests used a mobile vibrator that induced lateral excitation while traversing a track. The lateral accelerations were measured using a vibrator and selected sleepers. The tracks with a loosened ballast exhibited fewer differences between the vibrator and sleeper amplitudes, indicating reduced stability. Unsupported sleepers produced distinct "M-shaped" waveforms, owing to increased resistance at oversupported neighboring sleepers from redistributed loads. A key finding is the strong linear correlation between the lateral displacement of the vibrator in the dynamic tests and the lateral resistance in the static tests. This relationship enables an accurate estimation of lateral resistance from dynamic parameters, offering a scalable and efficient assessment method. The proposed dynamic method eliminates labor-intensive procedures and offers broader diagnostic coverage, making it effective for post-disaster inspections, maintenance planning, and routine evaluations that require rapid condition assessment
Bour's theorem for helicoidal surfaces with singularities
No full textIn this paper, by generalizing the techniques of Bour’s theorem, we prove that every generic cuspidal edge and, more generally, every generic n-type edge, which is invariant under a helicoidal motion in Euclidean 3-space admits non-trivial isometric deformations. As a corollary, several geometric invariants, such as the limiting normal curvature, the cusp-directional torsion, the higher order cuspidal curvature and the bias, are proved to be extrinsic invariants
Toward 100% fuel utilization in protonic ceramic fuel cells: modelling gas and current density distributions in a dead-end anode
No full textReaching 100% fuel utilization in protonic ceramic fuel cells (PCFCs) is advantageous for developing compact systems but remains challenging due to high overpotentials and anode material degradation caused by oxidation under low H2 partial pressure. We propose a dead-end anode design where the anode outlet is sealed, and a pre-filled H2O-H2 gas mixture ensures the required humidity for proton conductivity. The feasibility of 100% fuel utilization has been verified by a numerical model combining mass transfer (convection and diffusion) and charge transfer. Electrical performance is optimized by the pre-filled H2O mole fraction, height and length of the anode channel. Increasing the anode channel height and reducing its length significantly enhance performance of the dead-end-anode type PCFC, enabling a comparable current density–voltage performance at 100% fuel utilization to that of a relatively small coin-type configuration operating at 3% fuel utilization. Moreover, the combined effects of convection and diffusion of H2 and H2O gases help stabilize the anode gas partial pressures, thereby suppressing excessive local overpotentials. These findings, along with discussions on PCFC designs and potential applications, provide valuable insights for developing high-performance PCFCs with 100% fuel utilization for compact systems
Education effects of a seminar on defense-in-depth on individuals’ ability to identify risk-reduction measures in chemistry laboratories
No full textIt is essential to take a multilayered approach to risk reduction. This study examined whether the introduction of the concept of defense-in-depth, an approach used to achieve process safety, could allow university students and staff to propose improved, multilayered measures of the risk reduction for incidents that can occur in chemistry laboratories. A 10-minute seminar on defense-in-depth was conducted with university members. The participants completed questionnaires before and after the seminar. The collected data were analyzed with statistical methods and quantitative text analysis. After the seminar, the participants were able to propose a significantly larger number of safety measures per person for given incident scenarios than they could before the seminar. Furthermore, although before the seminar, participants’ attention was primarily focused on items that are easily imagined from the incident scenario, they proposed safety measures across all levels of defense-in-depth after the seminar. These findings are indisputable evidence that the seminar enabled participants to propose multilayered risk-reduction strategies. The results provide valuable insight into the development of more effective safety education programs and practices of risk reduction that can be implemented in chemistry laboratories
Factors Associated with Children's Problem Behaviors at the Time of Elementary School Enrollment: Focusing on Parenting Attitudes and Children's Self-Evaluation
Full text link
