Search

Currently, the number of visually impaired people is on the rise. There are many accidents involving the visually impaired while walking outdoors, and demand for walking support systems is high. In recent years, the number of stations with platform doors on the platform has been increasing, but there are still many stations that do not have such doors, which may cause accidents involving falls. In addition, visually impaired people cannot tell whether the steps in front of them are stairs or dangerous steps such as platforms. In this paper, we propose a system to detect upward and downward steps by combining depth images and CNN models. In particular, for downstairs steps, we distinguish between progressible steps such as stairs and dangerous steps such as platforms. In this research, we take depth images using a stereo camera that can obtain depth images, and create a data set. Using that dataset and CNNs, we will build a model that classifies four classes: “down_stairs”, “downstep”, “up_step”, and “plane”.

As the mileage of electric railways increases and deterioration over time progresses, the contact wire is more likely to wear, and the contact wire and the contact strip may separate. At that time, the contact wire is damaged and broken by arc generation and fusion and evaporation repetition. It is necessary to figure out the damage mechanism for stable supply of electricity and long-life and low cost of facility. However, it is difficult to clarify this phenomenon with experiment because it is a complicated physical phenomenon. Therefore, the elucidation of physical phenomenon using numerical analysis is required. The arc behavior and temperature distribution have been elucidated by the numeral method 1 - 5 . In this research, the simulation method considering electrode descent was developed in order to analyze the arc close to the actual phenomenon. As a result, the arc followed the descent of the contact strip and spread because of thermal diffusion as the distance between the electrodes increased. Therefore, the arc temperature distribution plays an important role to elucidate the disconnection of contact wire.

An arc lamp is a light source with high brightness and excellent color rendering emitted from an arc formed between electrodes, with high radiant power density and high energy plasma state. In industrial applications of arc discharge 1 2 , it is required to increase the luminance and to reach high temperature at high speed. One possible method is to increase the current value of the arc discharge, the Joule heating, and the temperature-dependent radiation 3 . It was confirmed that the increase in radiant power was different in the case of high temperature because of the gas in the arc lamp and in the case of a wide high temperature region 4 . However, when the input power increases, the energy efficiency decreases because the heat loss becomes larger. At the same time, in high-energy plasmas, high-order excitation and de-excitation occur, and UV light with high energy is generated. Therefore, to increase the energy efficiency, it is necessary to increase the amount of radiation at low temperature. With the change in air pressure under consideration of power efficiency, the objective is to propose the optimal air pressure for actual use in lamps. In this research, the arc temperature distribution of the short-arc lamp affected by changes in pressure was calculated using the 3-D electromagnetic thermal fluid simulation 5 .

Gas circuit breakers have a problem of thermal reignition. Therefore, to reduce the size of the arc quenching chamber while preventing thermal reignition and ensuring a reliable circuit breaker is need 1 . Therefore, to clarify the current interruption phenomenon under considering the cooling of the arc is needed. However, it is difficult to understand high-temperature arcs with experiments, so numerical analysis using simulations is being carried out in parallel 2 4 . In most of the simulations of circuit breakers, the arcs generated in a state where the interelectrode distance is constant and the poles are already open on the order of several centimeters have been analyzed. However, in an actual circuit breaker, the electrodes open with time, and the temperature volume of the arc also increases with time. In order to simulate such an opening behavior, it is difficult to handle the boundary between the electrode and the gas interface, where the physical quantity changes rapidly with time. In this research, the contribution of the opening speed for the reduction of the high temperature region in a double-flow circuit breaker is investigated using a three-dimensional electromagnetic thermo-fluid simulation.

Vacuum arc has been applied for valuable technology such as surface treatment of oxide layer 1 and arc ion plating (AIP). Although it has been required to elucidate the movement factor of cathode spots in vacuum arc, it is unclear because of the complexity of physics of cathode spot and difficulty of experimental elucidation. Intense cathode jet is produced at cathode spot because of evaporation from cathode spot caused by joule heating, emitted metal vapor is ionized. The behavior of these ions is considered to be a movement factor of cathode spot; however, the details have not been clarified. Therefore, numerical simulation of cathode jet distribution contributing to movement of cathode jet in vacuum arc was investigated in this research. In this research, three-dimensional electromagnetic thermal fluid simulation was developed 2 3 . Two-temperature between electron and heavy particle were considered in order to simulate plasma in vacuum arc 4 . The particle composition of thermal plasma near cathode was calculated by Saha’s equation. As a result, flow of neutral particles and ions from cathode jet contributing to movement of cathode spot is calculated, electromagnetic force and pressure gradient force acting on cathode jet was analyzed.

Recently, digital transformation is being promoted toward the establishment of next-generation power networks 1 2 , and in order to this achieve, it is necessary to collect analysis data under a wide variety of conditions. Circuit breakers are also required to be developed for the overall optimal solution of the power system 3 using monitoring and control technology, and it is expected that the efficiency of operations will be improved with the introduction of DX. The objective is to realize an air circuit breaker that suppresses switching surge voltage when shutting off, determines the opening speed according to the situation, and reduces the size and cost. Focusing on air circuit breakers, it is important to understand the basic characteristics of the arc generated between the electrode contacts in order to achieve miniaturization and high performance. However, few simulation analyzes of the opening speed of the air circuit breaker have been reported. In this research, the arc temperature distribution affected by the opening speed of the air circuit breaker was calculated using the 3-D electromagnetic thermal fluid simulation 4 . Specifically, a non-stationary analysis in which the movable contacts move diagonally over time is calculated.

The rapid increase in the development of machine learning systems in companies today presents new challenges that have not been faced in the development of traditional IT systems. Machine learning systems for human operations have a different set of challenges in comparison with Porter's connected smart products. The methodology for planning and developing such products has not yet been established, and it is mainly based on experience and intuition. In this paper, we define “exploitation” as the process of enhancing machine learning systems and summarize the issues in exploitation pattern based on several case studies. In the latter half of the paper, we propose a development process that minimizes machine learning problems on the exploitation of the systems. This is done by defining multiple uncertainties and proposing multiple development methods that are required for those uncertainties.

We propose a machine learning approach for power device design that consists of TCAD simulation and line Bayesian optimization. It has the advantage of multivariable optimization of field plate (FP) MOSFETs, and we thus discover an optimal structure for 100-V-class 8-step oxide FP-MOSFET. Compared with gradient FP-MOSFET, 8-step FP-MOSFET reduces the specific on-resistance (R ON A) by 20% and power losses by 6.2%–7.9%. Further, these losses can be reduced up to 11.4%–17.1% by considering parasitic charges in the objective function. We analyze the optimal 8-step structure, and discuss the reason why R ON A reduces.