• 제목/요약/키워드: probabilistic performance function

검색결과 92건 처리시간 0.026초

유/무성음 구분 및 이종적 특징 파라미터 결합을 이용한 화자인식 성능 개선 (Speaker Recognition Performance Improvement by Voiced/Unvoiced Classification and Heterogeneous Feature Combination)

  • 강지훈;정상배
    • 한국정보통신학회논문지
    • /
    • 제18권6호
    • /
    • pp.1294-1301
    • /
    • 2014
  • 본 논문에서는 화자 인식의 성능을 개선하기 위해서 유성음 및 무성음에 대한 별도의 확률분포 모델링을 사용하였다. 또한, 종래의 멜-주파수 캡스트럼 계수 이외에 유성음 구간에서 추가적으로 왜도, 첨도, 하모닉 대 잡음비 등을 추출하여 활용하였다. 화자 인식을 위한 스코어는 유성음 및 무성음 확률분포 모델에서 각각 구해지는데 전수 조사방식에 의해서 최적의 스코어 결합 가중치가 결정되었다. 제안된 방식의 화자인식기의 성능은 종래의 멜-주파수 캡스트럼 계수 및 화자당 하나의 혼합 가우시안 기반 확률분포 모델링을 사용한 방식과 비교되었으며 실험 결과 제안된 방식이 가우시안 혼합의 수가 낮아질수록 더 큰 성능 향상을 얻음을 알 수 있었다.

기계 구조의 강건 설계를 위한 최적화 기법의 개발 (Development of an Optimization Technique for Robust Design of Mechanical Structures)

  • 정도현;이병채
    • 대한기계학회논문집A
    • /
    • 제24권1호
    • /
    • pp.215-224
    • /
    • 2000
  • In order to reduce the variation effects of uncertainties in the engineering environments, new robust optimization method, which considers the uncertainties in design process, is proposed. Both design variables and system parameters are considered as random variables about their nominal values. To ensure the robustness of performance function, a new objective is set to minimize the variance of that function. Constraint variations are handled by introducing probability constraints. Probability constraints are solved by the advanced first order second moment (AFOSM) method based on the reliability theory. The proposed robust optimization method has an advantage that the second derivatives of the constraints are not required. The suggested method is examined by solving three examples and the results are compared with those for deterministic case and those available in literature.

신경회로망과 퍼지필터를 사용한 근전도신호의 기능변별에 관한 연구 (A Study on Function Discrimination for EMG Signals Using Neural Network and Fuzzy Filter)

  • 장영건;홍승홍
    • 대한의용생체공학회:의공학회지
    • /
    • 제15권3호
    • /
    • pp.355-364
    • /
    • 1994
  • The most important requirement for the controller of a prosthetic arm is that it has a high fidelity discriminator where the motion control may be performed open loop using EMG signals as a control source. Therefore, it is very effective method to reduce the influence of misclassification of classifier for the total system performance. This paper presents the new function discrimination method which combines MLP classifier and frizzy filter by stages for the requirement. The major advantage of MLP is a consistent learning capability for the easy adaptation to environments. The fuzzy filter uses all informations of MLP outputs and prior EMG activity informations which increase as the experience increases. That property is superior to one which uses maximum output of MLP in view of information amounts and quality. Simulation result shows that proposed method is superior to the probabilistic model, MLP model and the combined model of both in the respect of discrimination quaity.

  • PDF

A Simultaneous Design of TSK - Linguistic Fuzzy Models with Uncertain Fuzzy Output

  • Kwak, Keun-Chang;Kim, Dong-Hwa
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 제어로봇시스템학회 2005년도 ICCAS
    • /
    • pp.427-432
    • /
    • 2005
  • This paper is concerned with a simultaneous design of TSK (Takagi-Sugeno-Kang)-linguistic fuzzy models with uncertain model output and the computationally efficient representation. For this purpose, we use the fundamental idea of linguistic models introduced by Pedrycz and develop their comprehensive design framework. The design process consists of several main phases such as (a) the automatic generation of the linguistic contexts by probabilistic distribution using CDF (conditional density function) and PDF (probability density function) (b) performing context-based fuzzy clustering preserving homogeneity based on the concept of fuzzy granulation (c) augment of bias term to compensate bias error (d) combination of TSK and linguistic context in the consequent part. Finally, we contrast the performance of the enhanced models with other fuzzy models for automobile MPG predication data and coagulant dosing process in a water purification plant.

  • PDF

리니어 타입 크로스 롤러 가이드 공압 액추에이터의 신뢰성 평가에 관한 연구 (A Study on the Reliability Analysis for a Linear Type Pneumatic Actuator with Cross Roller Guide)

  • 신봉철;조명우;강성민;이수진;최진화
    • 한국정밀공학회지
    • /
    • 제23권5호
    • /
    • pp.184-189
    • /
    • 2006
  • This research presents the performance analysis of the linear type pneumatic actuators that are used in semi-conductor assembly line to transfer some product with high accuracy. To increase positioning and repetitive accuracies, a cross roller guide is implemented inside the pneumatic actuator. The finite element method is used to verify the force against working moments, and reliability analysis is performed to classify the breakdown cases. Also, reliability, failure rate, probability density function, and $B_{10}$ to life are estimated under the boundary of thrust or air leakage conditions. In this study, the failure probabilistic function of the pneumatic actuators is analyzed using Weibull distribution.

Probabilistic analysis of spectral displacement by NSA and NDA

  • Devandiran, P.;Kamatchi, P.;Rao, K. Balaji;Ravisankar, K.;Iyer, Nagesh R.
    • Earthquakes and Structures
    • /
    • 제5권4호
    • /
    • pp.439-459
    • /
    • 2013
  • Main objective of the present study is to determine the statistical properties and suitable probability distribution functions of spectral displacements from nonlinear static and nonlinear dynamic analysis within the frame work of Monte Carlo simulation for typical low rise and high rise RC framed buildings located in zone III and zone V and designed as per Indian seismic codes. Probabilistic analysis of spectral displacement is useful for strength assessment and loss estimation. To the author's knowledge, no study is reported in literature on comparison of spectral displacement including the uncertainties in capacity and demand in Indian context. In the present study, uncertainties in capacity of the building is modeled by choosing cross sectional dimensions of beams and columns, density and compressive strength of concrete, yield strength and elastic modulus of steel and, live load as random variables. Uncertainty in demand is modeled by choosing peak ground acceleration (PGA) as a random variable. Nonlinear static analysis (NSA) and nonlinear dynamic analysis (NDA) are carried out for typical low rise and high rise reinforced concrete framed buildings using IDARC 2D computer program with the random sample input parameters. Statistical properties are obtained for spectral displacements corresponding to performance point from NSA and maximum absolute roof displacement from NDA and suitable probability distribution functions viz., normal, Weibull, lognormal are examined for goodness-of-fit. From the hypothesis test for goodness-of-fit, lognormal function is found to be suitable to represent the statistical variation of spectral displacement obtained from NSA and NDA.

문장성분의 다양한 자질을 이용한 한국어 구문분석 모델 (Korean Parsing Model using Various Features of a Syntactic Object)

  • 박소영;김수홍;임해창
    • 정보처리학회논문지B
    • /
    • 제11B권6호
    • /
    • pp.743-748
    • /
    • 2004
  • 본 논문에서는 효과적인 구문 중의성 해결을 위해 문장성분의 구문자질, 기능자질, 내용자질, 크기자질을 활용하는 확률적 한국어 구문분석 모델을 제안한다. 그리고, 제안하는 구문분석 모델은 한국어의 부분자유어순과 생략현상을 잘 처리할 수 있도록 문법규칙을 이진형식으로 제한한다. 실험을 통해 제안하는 구문분석 모델의 성능을 각 자질조합별로 분석한다. 분석결과는 서로 다른 특징을 갖는 자질의 조합이 서로 유사한 특징을 갖는 자질의 조합보다 구문중의성 해결에 더 유용하다는 것을 보여준다. 또한, 단일자질인 기능자질이 내용자질과 크기자질의 조합보다 성능이 더 우수함을 알 수 있다.

Reliability-based assessment of damaged concrete buildings

  • Sakka, Zafer I.;Assakkaf, Ibrahim A.;Qazweeni, Jamal S.
    • Structural Engineering and Mechanics
    • /
    • 제65권6호
    • /
    • pp.751-760
    • /
    • 2018
  • Damages in concrete structures due to aging and other factors could be a serious and immense matter. Making the best selection of the most viable and practical repairing and strengthening techniques are relatively difficult tasks using traditional methods of structural analyses. This is due to the fact that the traditional methods used for assessing aging structure are not fully capable when considering the randomness in strength, loads and cost. This paper presents a reliability-based methodology for assessing reinforced concrete members. The methodology of this study is based on probabilistic analysis, using statistics of the random variables in the performance function equations. Principles of reliability updating are used in the assessment process, as new information is taken into account and combined with prior probabilistic models. The methodology can result in a reliability index ${\beta}$ that can be used to assess the structural component by comparing its value with a standard value. In addition, these methods result in partial safety factor values that can be used for the purpose of strengthening the R/C elements of the existing structure. Calculations and computations of the reliability indices and the partial safety factors values are conducted using the First-order Reliability Method and Monte Carlo simulation.

원자력발전소 직류 전력계통의 충전기 신뢰도 향상방안 연구 (A Study on Battery Charger Reliability Improvement of Nuclear Power Plants DC Distribution System)

  • 임혁순;김두현
    • 한국안전학회지
    • /
    • 제25권2호
    • /
    • pp.24-28
    • /
    • 2010
  • The nuclear power Plant onsite AC electrical power sources are required to supply power to the engineering safety facility buses if the offsite power source is lost. Typically, Diesel Generators are used as the onsite power source. The 125 VAC buses are part of the onsite Class 1E AC and DC electrical power distribution system. The DC power distribution system ensure the availability of DC electrical power for system required to shutdown the reactor and maintain it in a safety condition after an anticipated operational occurrence or a postulated Design Base Accident. Recently, onsite DC power supply system trip occurs the loss of system function. To obtain the performance such as reliability and availability, we analyzed the cause of battery charger trip and described the improvement of DC power supply system reliability. Finally, we provide reliability performance criteria of charger in order to ensure the probabilistic goals for the safety of the nuclear power plants.

고장 데이터의 확률 분석을 적용한 항공기 시스템 안전성 평가 수행 방안 연구 (A Study on the Implementation of Aircraft System Safety Assessment using Probabilistic Analysis of Failure Data)

  • 유승우;김인걸
    • 항공우주시스템공학회지
    • /
    • 제14권spc호
    • /
    • pp.31-38
    • /
    • 2020
  • 항공기의 개발 및 인증 과정에서 중요성이 강조되고 있는 시스템 안전성 평가는 고장 또는 결함으로 인해 발생할 수 있는 위험을 사전에 식별하고, 이에 대한 영향성을 판단하여 요구되는 수준의 안전성을 확보하기 위해 적용하는 시스템 엔지니어링 기법이다. 항공기 시스템이 복잡해지고, 통합적인 기능 및 성능이 요구되면서 항공기 개발 초기부터 적절한 안전성 목표를 수립하고, 체계적인 평가를 통해 이에 대한 검증 및 후속 조치를 이행하여야 한다. 본 논문에서는 고장 데이터의 확률 분석을 수행하는 방안을 제시하여 항공기의 전체 수명주기에 걸쳐 리스크를 진단하고 효율적인 안전성 평가를 수행하기 위한 방법론을 제시하였고, 항공기 인증 및 안전성 확보를 위해 고려해야 할 사항을 제시하였다.