http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
LSTM 순환 신경망을 이용한 초음파 도플러 신호의 음성 패러미터 추정
주형길,이기승,Joo, Hyeong-Kil,Lee, Ki-Seung 한국음향학회 2019 韓國音響學會誌 Vol.38 No.4
본 논문에서는 입 주변에 방사한 초음파 신호가 반사되어 돌아올 때 발생하는 초음파 도플러 신호를 LSTM(Long Short Term Memory) 순환 신경망 (Recurrent Neural Networks, RNN)을 이용해 음성 패러미터를 추정하는 방법을 소개하고 다층 퍼셉트론 (Multi-Layer Perceptrons, MLP) 신경망을 이용한 방법과 성능 비교를 하였다. 본 논문에서는 LSTM 순환 신경망을 이용해 초음파 도플러 신호로부터 음성 신호의 푸리에 변환 계수를 추정하였다. LSTM 순환 신경망을 학습하기 위한 입력 및 기준값으로 초음파 도플러 신호와 음성 신호로부터 각각 추출된 멜 주파수 대역별 에너지 로그값과 푸리에 변환 계수가 사용되었다. 테스트 데이터를 이용한 실험을 통해 LSTM 순환 신경망과 MLP의 성능을 평가, 비교하였고 척도로는 평균 제곱근 오차(Root Mean Squared Error, RMSE)가 사용되었다.각 실험의 RMSE는 각각 0.5810, 0.7380로 나타났다. 약 0.1570 차이로 LSTM 순환 신경망을 이용한 방법의 성능 우세한 것으로 확인되었다. In this paper, a method of estimating speech parameters for ultrasonic Doppler signals reflected from the articulatory muscles using LSTM (Long Short Term Memory) RNN (Recurrent Neural Networks) was introduced and compared with the method using MLP (Multi-Layer Perceptrons). LSTM RNN were used to estimate the Fourier transform coefficients of speech signals from the ultrasonic Doppler signals. The log energy value of the Mel frequency band and the Fourier transform coefficients, which were extracted respectively from the ultrasonic Doppler signal and the speech signal, were used as the input and reference for training LSTM RNN. The performance of LSTM RNN and MLP was evaluated and compared by experiments using test data, and the RMSE (Root Mean Squared Error) was used as a measure. The RMSE of each experiment was 0.5810 and 0.7380, respectively. The difference was about 0.1570, so that it confirmed that the performance of the method using the LSTM RNN was better.
단일전류센서를 이용한 교류전동기 구동에서 전동기 상수동정과 그 오차
주형길(Hyeong-Gil Joo),신휘범(Hwi-Beom Shin),윤명중(Myung-Joong Youn) 전력전자학회 2000 전력전자학술대회 논문집 Vol.2000 No.11
An estimation scheme is used for solving two practical problems of the single current sensor technique. To improve the effect of parameter uncertainties, the method that identifies motor parameters for AC motor drive systems using the single current sensor technique is presented. And, the parameter identification error and its cause are examined. It gives good performances for identifying parameters. and reconstructing phase currents.
스누버 전류를 고려한 개선된 SVPWM 인버터를 이용한 상전류센서없는 전동기 구동
주형길(Hyeong Gil Joo),신휘범(Hwi Beom Shin),안희욱(Hee Wook Ahn),윤명중(Myung Joong Youn) 전력전자학회 1999 전력전자학술대회 논문집 Vol.1999 No.7
The single sensor technique reconstructing phase currents from the dc-link current without phase current sensors is proposed. When the duration of active vector is too short for the snubber current to reduce, the dc-link current including the snubber current gives large detection error The solution is presented by analyzing the snubber current and mollifying the switching sequences. This scheme is simple, requires only one sampling a period and has good results for detecting the phase currents
정전압 변압기의 회로정수에 따른 특성 변화에 관한 연구
주형길(Hyeong-Gil Joo),한재원(Jae-Won Han) 전력전자학회 2001 전력전자학술대회 논문집 Vol.2001 No.7
The constant voltage transformer(CVT) is a voltage regulator that has a high reliability, a good short circuit characteristic, a high input power factor, and a relative immunity to lightning strike. Moreover, this has a good line regulation without a voltage feedback control. In this paper, the characteristics of CVT for the changes of input voltage and other circuit parameters are investigated.
백인철,주형길 전력전자학회 2005 JOURNAL OF POWER ELECTRONICS Vol.5 No.1
A simple fault correction method for rotor position detection of a brushless DC(BLDC) motor with trapezoidal back EMF(electromotive force) using a Hall effect latch unit is presented. The reason why the Hall effect latch unit does not operate properly during the startup of a BLDC motor is thoroughly explained. To solve this problem, a simple code change method and its hardware implementation issues are proposed and discussed.
장두원(Doo-Won Chang),주형길(Hyeong-Gil Joo) 전력전자학회 2008 전력전자학술대회 논문집 Vol.- No.-
최근 전원부의 소형화 경향에 따라 에너지 변환 소자인 변압기의 소형 경량화 요구가 가속화 되고 있다. 전원부의 고효율소형화를 위해서 평면 변압기가 대두되고 있다. 이에 따라 본 논문에서는 8[W]급 플라이-백(Flyback) 평면변압기의 PCB(Printed Circuit Board) 패턴을 FERROXCUBE 사의 평면 변압기 설계 방법을 따라 개념 설계를 하였다. 120[kHz]의 스위칭 주파수에서 동작하는 평면 변압기의 자계분포(magnetic intensity), 와전류(eddy current), 전계 분포(electric intensity)를 유한 요소법(Finite Element Method)을 이용하여 분석하였다.
3.3㎸, 15㎸A급 3-레벨 NPC PWM 컨버터의 중성점 전압제어에 관한 연구
안준선(Joonseon Ahn),주형길(Hyeong-Gil Joo),서명석(Myung-Seok Seo),문승일(Seung-Il Moon),김주용(Ju-Yong Kim),송일근(Il-Geun Song) 한국조명·전기설비학회 2011 한국조명·전기설비학회 학술대회논문집 Vol.2011 No.5월
Recently, many researchers has been focused the “Smart Grid” system because the market of replacement of old distribution system has been increasing. The solid-state transformer is one of main component of “Smart Grid” system, it is important to develop the transformer. In this paper, 3.3[㎸] 15[㎸A] solid-state transformer has developed. Three-level diode clamped PWM converter-inverter was used in the transformer, therefore the voltage balancing algorithm between DC link capacitors also applied. We verified the solid-state transformer with computer simulation and experiment, voltage balancing was kept well