• Title/Summary/Keyword: 정적 천이 속도

Search Result 9, Processing Time 0.022 seconds

Variable Time-Scale Modification of Speech Using Trasient Information (천이구간 정보를 이용한 음성의 가변적인 시간축 변환)

  • Lee, Sung-Joo;Kim, Hee-Dong;Kim, Hyung-Soon
    • Journal of the Korean Institute of Telematics and Electronics S
    • /
    • v.35S no.6
    • /
    • pp.147-155
    • /
    • 1998
  • Conventional time-scale modification methods have the problem that as the modification rate gets higher the time-scale modified speech signal becomes less intelligible, because they ignore the effect of articulation rate on speech characteristics. In this paper, we propose a variable time-scale modification method based on the knowledge that the timing information of transient portions of a speech signal plays an important role in speech perception. After identifying steady protions only. The result of subjective preference test indicates that the proposed method produces performance superior to that of the conventional SOLA method.

  • PDF

A Study on Combustion Characteristics of the Methane-Hydrogen Lean Mixture by Using Multiple Spark Capacity Discharge in a CVCC (II) (반복점화장치 사용시 정적연소실내 메탄-수소 희박혼합기의 연소특성 연구(II))

  • Kim Bong-Seock
    • Journal of Energy Engineering
    • /
    • v.13 no.4
    • /
    • pp.311-318
    • /
    • 2004
  • In the present study, the combustion characteristics of methane and hydrogen-supplemented methane as alternative fuels for automotive vehicles were investigated at various hydrogen substitution rate, ignition position and ignition methods in a CVCC. The main results obtained from the study can be summarized as follow. In case of center ignition and neat methane-air mixture, the flame propagation processes are propagated with an elliptical shape, but they are changed an instable elliptical shape flame with very regular cells and higher velocity by increasing the hydrogen supplement rate. In case of side, 0.5R ignition and neat methane-air mixture, the flame propagation processes are propagated with an instable elliptical shape flame, but they are changed from an instable elliptical shape to wedge shape flame with very irregular cells and higher velocity by increasing the hydrogen supplement rate. Although the flame propagation shape with ignition position and ignition devices was not differ, the flame area of MSCDI device was a little larger than it of CDI device at the same time.

Generating LOTOS Specifications from UML Static Structure Diagrams (UML 정적구조 다이아그램으로부터 LOTOS 명세 생성)

  • Kim, Cheol-Hong;Ahn, Yu-Whoan;Lee, Won-Chun
    • The Transactions of the Korea Information Processing Society
    • /
    • v.6 no.12
    • /
    • pp.3500-3513
    • /
    • 1999
  • It is recognized that object-oriented methods and formal methods are two different main streams that will influence on the future direction of software engineering. A merging effort on these two technologies, named "a formal approach on system specifications using object-oriented methods" emerges rapidly and produces remarkable research results LOTOS is well-suited to an object-based approach. However, to provide a full object-oriented approach, we need to model generalization (i.e. inheritance and polymorphism). Most authors who have examined this topic have proposed extensions to LOTOS. As an extension of such an effort, this paper proposes a method that generates LOTOS specification from static structure diagrams in UML.

  • PDF

Dynamic Filter Pruning for Compression of Deep Neural Network. (동적 필터 프루닝 기법을 이용한 심층 신경망 압축)

  • Cho, InCheon;Bae, SungHo
    • Proceedings of the Korean Society of Broadcast Engineers Conference
    • /
    • 2020.07a
    • /
    • pp.675-679
    • /
    • 2020
  • 최근 이미지 분류의 성능 향상을 위해 깊은 레이어와 넓은 채널을 가지는 모델들이 제안되어져 왔다. 높은 분류 정확도를 보이는 모델을 제안하는 것은 과한 컴퓨팅 파워와 계산시간을 요구한다. 본 논문에서는 이미지 분류 기법에서 사용되는 딥 뉴럴 네트워크 모델에 있어, 프루닝 방법을 통해 상대적으로 불필요한 가중치를 제거함과 동시에 분류 정확도 하락을 최소로 하는 동적 필터 프루닝 방법을 제시한다. 원샷 프루닝 기법, 정적 필터 프루닝 기법과 다르게 제거된 가중치에 대해서 소생 기회를 제공함으로써 더 좋은 성능을 보인다. 또한, 재학습이 필요하지 않기 때문에 빠른 계산 속도와 적은 컴퓨팅 파워를 보장한다. ResNet20 에서 CIFAR10 데이터셋에 대하여 실험한 결과 약 50%의 압축률에도 88.74%의 분류 정확도를 보였다.

  • PDF

A Study on Three-dimensional Dynamic Analysis of a Towing Cable for Marine Survey Instruments (해양탐사장비 예인케이블의 3차원 동적해석에 관한 연구)

  • 정동호;김종규;박한일
    • The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
    • /
    • v.8 no.2
    • /
    • pp.203-209
    • /
    • 2003
  • In this study, the configuration and tension of a towing cable for side-scan sonar are predicted in an ambient flow and at an unsteady towing condition. The governing equation of three-dimensional dynamic analysis for a flexible cable is solved using a finite difference method. We successfully predict the configuration and tension of a side-scan sonar and designed the towing system. It is found in static analyses that the side-scan sonar must be towed to keep a its stable depth at a reasonable speed. The study also reveals in the transient analyses that the dominant component affecting the top tension is the tangential drag force for the larger towing speed than the critical speed, and the soft weight of a towed instrument for the smaller towing speed than. It should be maneuvered for a towing vessel with good consideration for the impact effect in a cable due to tension peak when a towing speed is suddenly increase. The developed program can be applicable for three-dimensional dynamic analysis of a towing system for various marine survey instruments.

A Dry Friction Model to Realize Stick for Simulation of the System with Friction and Accuracy Verification of the Friction Model (마찰력이 작용하는 동적 시스템의 점착 구현을 위한 마찰모델 제안 및 정확성 검증)

  • Choi, Chan-Kyu;Yoo, Hong-Hee
    • Transactions of the Korean Society for Noise and Vibration Engineering
    • /
    • v.22 no.8
    • /
    • pp.748-755
    • /
    • 2012
  • Friction causes self-excited vibration, stick-slip vibration and any other friction-induced phenomena. That kinds of vibrations cause chatter and squeal. In order to predict such vibrations accurately, employing an accurate friction model is very important because a dynamic behavior of a system with friction is dominantly governed by a friction model. A Coulomb friction model is the most widely known model. Coulomb friction model is useful model to obtain analytical solutions of the system with friction and the model gives relatively good simulation result. However, defining a friction force at a stick state in simulation is hard because of the characteristic itself and a Coulomb friction model is discontinuous function between a static and a dynamic friction coefficient. Therefore, applying the Coulomb friction model to a simulation is not appropriate. In order to resolve these problems, an approximated Coulomb friction model was developed using simple and continuous function. However, an approximated Coulomb friction model cannot realize stick. Therefore, an approximated Coulomb friction model cannot describe friction phenomena accurately. In order to analyze a friction phenomenon accurately, a friction model for a simulation was proposed in this paper. A proposed friction model realizes stick and gives reasonably good results compared to results obtained by the simulation employing an approximated Coulomb friction model. Accuracy of a proposed friction model was verified by comparing experimental results.

Behavior of the Crack Initiation, Transition and Fatigue Crack Growth of Rail Steel (레일강의 균열발생·천이 및 피로균열진전거동)

  • Lee, Jong Sun;Kang, Ki Weon;Choi, Rin;Kim, Jung Kyu
    • Journal of Korean Society of Steel Construction
    • /
    • v.11 no.1 s.38
    • /
    • pp.33-42
    • /
    • 1999
  • In the present study, crack initiation criteria, static failure and tensile mode fatigue behavior for a rail steel are evaluated to assure the railway vehicle's safety. The transverse fissure, which is the most critical damage in the rail, is initiated by the maximum shear stress and its location is subsurface. In addition, the possibility of transition from the shear mode to the mixed mode increases with increasing the length of subsurface crack. Because of the brittleness by the welding, the fracture toughness of the welded part is lower than of the base metal. For low ${\Delta}K$, the stage II fatigue crack growth rates of the welded part is slower than of the base metal but, for high ${\Delta}K$, this different behavior for fatigue crack growth rate is nearly diminished. These trends are more remarkable for low stress ratio, R=0.1. It is believed that this behavior is caused by the change of the microstructure which that of the welded part is coarser than of base metal.

  • PDF

Steady-state and Transient Performance Simulation and Limit Control for Compressor Surge and Turbine Over-temperature of Turboprop Engine (PT6A-62) (터보프롭 엔진(PT6A-62)의 동.정적 성능모사와 압축기 서지 및 터빈 자온 제어연구)

  • 공창덕;기자영;강명철
    • Journal of the Korean Society of Propulsion Engineers
    • /
    • v.6 no.2
    • /
    • pp.53-63
    • /
    • 2002
  • The steady-state and transient performance simulation program for a turboprop engine(PT6A-62) was developed. Specially this program included some algorithms, such as flat-rated behaviors in performance and limit control algorithms to prevent the compressor surge and the compressor-turbine inlet limit temperature overshoot. In order to minimize analysis errors, on interpolation method in component characteristics using matching errors and specific heat and specific heat ratio, which are functions of temperatures were used. The developed steady state performance analysis program can handle various conditions such as altitude, bleed extraction, inlet temperature and pressure and part throttle, and the transient performance analysis program incorporated a general mode for transient simulation and a control mode for prevention of the compressor surge and the turbine inlet limit temperature overshoot.

Damage Estimation for Offshore Tubular Members Under Quasi-Static Loading (준정적하중(準靜的荷重)을 받는 해양구조물(海洋構造物)의 원통부재(圓筒部材)에 대한 손상예측(損傷豫測))

  • Paik, Jeom-K.;Shin, Byung-C.;Kim, Chang-Y.
    • Bulletin of the Society of Naval Architects of Korea
    • /
    • v.26 no.4
    • /
    • pp.81-93
    • /
    • 1989
  • The present study attempts to develop the theoretical model for the damage estimation of offshore tubular members which are subjected to the accidental impact loads due to collision, falling objects and so on. For the reasons of the simplicity of the problem being considered, however, this paper postulates that the accidental load can be approximated to be the quasi-static one, in which dynamic effects are negelcted. Based upon the theoretical and experimental results which are obtained from the present study as well as the existing literature, the load-displacement relations taking the interaction effect between the local denting and the global bending deformation into account are presented in the explicit form when the concentrated lateral load acts on the tubular member whose end condition is supposed to be rotation ally free and axially restrained, in which membrane forces develop. Thus, the practical estimation of damage deformation for the local denting and the global bending damage of tubular members against the accidental loads is possible and also the collision absorption capability of the member can be calculated by performing the integration of the area below the given load-displacement curves, provided that all the energy is dissipated to the deforming the member itself.

  • PDF