1. はじめに

「電気関係学会関西連合大会」特集号は「関西からの情報発信，連合大会の活性化，論文誌Cへの投稿の促進」を目的として，平成29年電気関係学会関西連合大会で発表された優れた研究成果を論文としてまとめて掲載するものである。電気学会論文誌Cの編修委員会のご理解とご協力によ

Characteristics of flashover voltage were investigated for some electrode configurations (gap on borosilicate glass, gap on ZnO, air gap). And negative steep pulse voltage was applied to the gap on solid dielectric plate with the gap lengths of 5 ～ 40 µm. As a result, it found that the flashover voltage in each electrode configuration depends on the gap length. At the gap lengths of 5 ～ 20 µm, no significant difference among the investigated electrode configurations at flashover voltage was observed. Moreover, at 5 ～ 10 µm, it was close to the Paschen curve. At the gap lengths of 20 ～ 40 µm, the flashover voltage varies depending on the electrode configuration, and the descending order are the air gap, the glass gap, and the ZnO gap. It is suggested the property that ZnO's work function is low make the flashover voltage reduction.

The conjugated system of polymer has an important role on the photovoltaic properties of a polymer bulk heterojunction (BHJ) solar cell. In this work, we focused on the impact of π-conjugation length of poly(3-hexylthiophene) (P3HT) on the power conversion efficiency of P3HT:phenyl-C₆₁-butyric-acid methyl ester (PCBM) BHJ solar cell. The π-conjugation length of P3HT is systematically controlled through partial bromination at the 4-position of thiophene ring, which elucidated the correlation between the degree of P3HT bromination and the photovoltaic properties of Br-P3HT:PCBM BHJ solar cells. The energy conversion efficiency of the unbrominated (pristine) P3HT:PCBM solar cell is 2.3%, whereas that of Br-P3HT:PCBM solar cell with approximately 7% bromination falls to 1.4%. The present results suggest that the π-conjugation length of about 14 thiophene rings is insufficient for achieving efficient charge transport.

This paper describes an analyzing method of the optical directional couplers composed of nonlinear waveguides with perturbed structure parameters. The dyadic Green-function is applied to solving the nonlinear coupled-mode equations for the purpose of easily understanding physical meanings of the power transfer characteristics in the perturbed structures. Advantages of the formula using the dyadic Green-function are shown in optimizing parameter designs of the nonlinear coupler devices, and designed operation characteristics with designed parameters are also performed in a power divider working in the wide-range input light-power, light switch operating by the input light power, and light power filter in which light output is gotten in the desired narrow input power range.

In recently years, the ground capacitance has been increased due to the development of switching elements and others. It cannot be ignored that the capacitive leakage current generated by them. Because of them, it is difficult that the resistive leakage current which is generated by insulation fault and others has been detected. The existing leakage circuit breaker which is used by zero phase current transformer (ZCT) cuts off a circuit even when the capacitive leakage current only flows. So it is very important that we only detect the resistive leakage current. The magnetic thin film functional sensors we produced is compact and can measure real power from apparent power with high accuracy. Therefore, we experimented detecting the resistive leak current from all leakage current by it. The result is that it can detect the resistive leak current. In this paper, we report that the magnetic thin film functional sensors we produced can use the key device of leakage circuit breaker.

A multi-level inverter has potential advantages in high blocking voltages and reduced harmonics. However, it has a technically challenging subject that the DC input voltages must be balanced. This paper proposes and investigates a new double-mapping-based space-vector modulation (SVM) method to solve the subject without any additional circuit. The proposed method consists of two steps. One is the SV map data generation for the five-level NPC inverter based on a general principle to utilize redundant SVs. The other is the optimal selection of the SV pattern which balances the the DC input voltages based on the double-mapping-based SVM method which has two different kind of control schemes for low and high modulation indexes. The proposed method is investigated and validated through simulated results by a digital control system.

The aim of this study is to capture the tongue movement in an infant during sucking. We developed a glove type sensor device using six small pressure sensors and measured the tongue pressure during sucking. Furthermore, we estimated the tongue pressure distribution. As a result, the waveform considered to be due to peristaltic-like movements of the tongue were captured. From the measurement results of more than 100 infants, we clarified that the sucking ability can be evaluated at two points: ①the presence of force necessary for sucking, and ②the presence of peristaltic-like movement. In this study, we propose a method to quantitatively evaluate sucking ability and suggest the possibility of application to household equipment.

We have developed an optical transmitter that tracks an optical repeater in the underwater video transport system. It can detects the misalignment of visible light with single sine wave. We examined its effectiveness of its detective performance by using newly developed four optical modules. The detection performance achieved in the experiment well agreed with the theoretical value. We confirmed the feasibility of the beam tracking method.

In the field of geological survey, a technique called electrical exploration method is widespread, and survey of underground structure can be realized using multi electrodes arranged on the ground surface. This research is aimed to realize EIT by small electrode sheet using the principle of electrical exploration method. In this paper, impedance distribution of a water tank with conductive gel that can easily prepared was measured by the electrical exploration method, and the effectiveness and problem of bioimpedance measurement was discussed.

In recent years, Near-Infrared Spectroscopy various kinds are developed and used in some purposes. In this study we measure cerebral blood volume with dementia patients and health elderly controls using Near-Infrared Spectroscopy by two wavelengths. Category fluency task was used as performing of blood measurement. Oxy-hemoglobin and total-hemoglobin were evaluated as index of blood volume. 37 dementia patients (age 82.5 ± 5.5, 11 male, 26 female) and 20 health elderly controls (age 74.2 ± 5.6, 2 male, 18 female) were instructed to produce a noun belonging to the category of vegetables while measuring cerebral blood volume. In health elderly controls, oxy-hemoglobin was significantly greater at task than rest in all channels. On the other hand, oxy-hemoglobin with dementia patients wasn't showed increasing in all channels. In addition, Total-hemoglobin of dementia patients increased while category fluency task in right area.

A gradient based active noise control system using the frequency domain SSCF (Square Sum of Correlation Function) adaptive algorithm without a model of the secondary path is proposed. The cost function of this adaptive system is a square sum of the cross-correlation coefficient of the input noise and the output error signal rather than mean square of the output error signal. A observable joint statistics of two signals instead of the estimation of a unobservable signal is considered. The spectrum detection method of the noise source and the influence of the detection error on adaptive algorithm are investigated, as a result the stability of the algorithm was proved. Since this algorithm is a normalized algorithm in the frequency domain, the adaptive parameters independently converge to the optimum value. Numerical verifications are also presented by using non-stationary noise source observed in passenger car and the actually measured transfer functions of the acoustic path.

In this paper, we aim to find out the energy efficiency of an individual lighting control system using image-based motion sensors, which recognize occupants with image recognition technologies, by comparing a lighting control system using conventional passive infrared sensors (PIR sensors). We evaluated the actual energy consumption by the lighting control system using the two types of sensors over 12 weeks in 2 office environments and simulated energy consumption based on the record of the sensor values. Our result suggests that the individual lighting control system using the image-based sensors have reduced the energy consumptions by more than 17.3% (0.83 kWh) per day on median. In addition, we analyzed the advantages and disadvantages of the individual lighting control system based on image-based sensor in more detail. One of the disadvantages was misdetection of occupants, which may increase energy consumption. From estimation based on simulation, we consider that the effect of misdetection may be reduced by time-scheduling control.

Immunity tests are a kind of electromagnetic compatibility test. Tests examine the failure of electronic equipment when electromagnetic noise is applied. Immunity tests take long time, and failures in test are detected by means of visual observation in most cases. Therefore tests are a burden to examiners. With the aim of reduction of the burden, we have developed “Failure detection assist system for immunity tests”. In the case of a rotator such as motor, it is the failure that rotation speed is varied unintentionally. In this paper, we propose a rotation speed detection technique, which employs a camera and a marker, for immunity tests. We experimented to detect the variation of rotation speed, and confirmed that the proposed technique can discriminate between steady state and unsteady state about motor rotation.

Optimization problems under supply and demand balance are important in energy management such as electric and heat management. The authors have proposed a fast solution method for the optimization problems based on optimality conditions and formula manipulation algorithm. In this paper, we extend the original optimization problem by introducing weighting factors and offset term for supply-amounts. By this extension, we can deal with wider range problems such as optimal fuel distribution problems for steam boilers. We also present an extended fast solution method for the extended problem.

This paper considers a functional electrical stimulation (FES) cycling with cadence tracking control for rehabilitation of hemiparesis. A recumbent trike and rider with four muscle groups is modeled as an Euler Lagrange system based on a pedaling force direction. We propose cadence tracking control considering effects of antagonistic bi-articular muscles to stimulate multiple muscle groups continuously at all angles during pedaling motion. Stability of the developed controller is analyzed through Lyapunov-based methods. Experimental results in healthy participants are shown to confirm the validity of the proposed method and the potential for the rehabilitation of hemiparesis patients.

Recently, the spread of Web cameras has facilitated video-conferencing as a means of communication. Since a Web camera is usually located outside the display while the user looks at his/her partner in the display, there is a problem that they cannot establish eye contact with each other. Various methods have been proposed to solve this problem, but most of them required specific sensors, and robust and accurate feature extraction for various appearances to densely associate facial feature points of the images. In this paper, we propose a method that transforms the eye areas to synthesize eye contact using a single camera that is commonly implemented in laptop computers. Concretely, we implemented a system which transforms the user's eye areas in an image to his/her eye image with a straight gaze to the camera only when the user's gaze falls in a range that the partner would perceive eye contact. We then confirmed the effectiveness of the proposed method in terms of accuracy of the gaze area classification, subjective evaluation score for eye contact, and accuracy of eye contact through experiments.

This paper proposes an impact noise suppression method using a speech spectral phase estimator. We firstly estimate a speech and a noise spectral amplitudes in a zero-phase (ZP) domain. Next, we estimate the noise spectral phase as a slope of an observed spectral phase. Subtracting the estimated noise spectral amplitude with phase from the observed spectrum, and taking the spectral phase of it, we have the estimated speech spectral phase. Taking the IDFT of the estimated speech spectral amplitude with phase gives the estimated speech signal in time domain. When the analysis frame does not include the noise, we should not perform the noise suppression for avoiding any speech degradation. Hence, we introduce the Teager-Kaiser energy operator to detect the noise-existence frame with a high accuracy. Simulation results show that the proposed impact noise suppressor improves SNR, PESQ, and STOI scores in comparison to conventional methods.

Many digital pictures are used as evidence in criminal investigation. It is very important to check whether they are doctored or not. This paper proposes a method for detecting doctored region in JPEG image using a convolutional neural network (CNN). In the proposed method, DCT coefficients are input to the CNN. Its output is a binary segmented image in which doctored and non-doctored regions are shown with white and black pixels, respectively. In our experiment, 45 types of CNN models were created and compared. The detection accuracy of the best model achieved 0.63 in terms of F-measure, which is larger about 2.3 times than that of our preliminary method based on support vector machine (SVM).

This paper is associated with an auto-tuning stochastic resonance approach for image enhancement on illumination variant images. This new process is being developed based on our previous works related to image enhancement by using stochastic resonance (SR) by using the manual tuning process. The process was performed by adding the random noise and threshold 0 in an image. The process works properly in the dark and very low contrast images as well as bright images based and mixed illumination variant images. This system works on the images with the mixture of darkness and brightness. We also present the idea of auto-tuning of the iteration with random noise and threshold value 0 by using the process related to the histogram calculation or mean and median. We performed various experiments on various types of images including human faces under different conditions and confirmed the effectiveness of our image enhancement technique.

Since automotive technologies are dramatically improved nowadays, HMI development for in-vehicle system becomes complicated to design for many functions. Therefore, man-months of HMI development are increased. In order to solve the issue, Model-based HMI development method is proposed. HMI model is defined to represent HMI design specifications as formatted information. Conversion method from an image editor and a layout editor to HMI design specifications via the HMI model is illustrated. As a result of evaluation, peak time of development using proposed method was shorter than conventional method and efficiency of publishing HMI design specifications to HMI programmers was improved.

Network study is one of the very important subjects to be taught not only in the higher education of science and technology fields but also in other introductory courses of the lower education of the world. Kagawa university had developed an educational tool for the sake of studying network topology and Internet Protocol, so that it wanted to provide its tool widely its tool in order for any potential learners of the world to utilize network study, but it must overcome the problem of language barrier for foreign students to take an advantage of the tool in their mother languages. This time, such a tool has been improved into a world-widely available e-Learning system, which can equip multilingual supports and explanations for Japanese learners as well as French/English/Spanish/Portuguese spoken learners. Thanks to these multilingual supports and explanations, a newly improved system will bring a very useful educational environment to learners of the world and will let them manipulate such a newly modified e-Learning system of network study effectively and efficiently. The modified e-Learning system has been well evaluated through performing multiple questionnaires and receiving comments from foreign learners.

In this paper, we propose the 1+n/k frequency divider based on the multi-phase clock for the purpose of improving the accuracy of the dividing ratio changeable all digital PLL. The division ratio of this circuit is not limited to n ≤ ⌊k/2⌋ (⌊x⌋ is the largest integer not exceeding x)。 Also, it is confirmed by simulation that the intended operation can be obtained also for the step response of the division ratio.

This study propose a training system that promotes isotonic exercise of the styloglossus muscle for tongue rehabilitation. The proposed system uses a DC motor to operate a tongue depressor that press the tongue to induce isotonic exercise. Electromyography signals measured from the gums indicate the isotonic contraction of the styloglossus muscle during the exercise.