• 제목/요약/키워드: Convex Function

검색결과 446건 처리시간 0.026초

가스터빈 엔진의 복합 결함 진단을 위한 SVM과 MLP의 성능 비교 (A Performance Comparison of SVM and MLP for Multiple Defect Diagnosis of Gas Turbine Engine)

  • 박준철;노태성;최동환
    • 한국추진공학회:학술대회논문집
    • /
    • 한국추진공학회 2005년도 제25회 추계학술대회논문집
    • /
    • pp.158-161
    • /
    • 2005
  • 본 연구에서는 Support Vector Machine (SVM)을 이용하여 가스 터빈 엔진의 결함 진단을 시도하였다. SVM은 벡터 공간에서 임의의 비선형 경계인 Hyperplane을 찾아 두 개의 집합을 분류하는 방법으로 수학적으로 최적의 해를 찾을 수 있다고 알려져 있다. 이러한 이진 분류용 SVM을 다층으로 결합하여 가스 터빈의 결함을 정량적으로 판단해 내는 방법을 제안하였으며 기존의 Multi Layer Perceptron(MLP)보다 빠르고 신뢰성 있는 진단 결과를 보여주었음을 확인하였다.

  • PDF

점탄성 감쇠기가 설치된 실물크기 5층 철골건물의 진동실험 (Dynamic Experiment of a Full-Scale Five-story Steel Building with Viscoelastic Dampers)

  • 민경원;이영철;이상현;박민규;김두훈;박진일;정정교
    • 한국지진공학회:학술대회논문집
    • /
    • 한국지진공학회 2002년도 추계 학술발표회 논문집
    • /
    • pp.239-246
    • /
    • 2002
  • Viscoelastic dampers are known effective devices for response reduction under earthquakes and winds. This study addresses how to design the optimum viscoelastic dampers installed at the full scale five-story steel building and novel approach to carry out the experimental work to verify the damper performance. First, an exciter of hybrid mass-type actuator is designed, which can move the building and its mathematical model is derived. The integrated system of building-actuator is experimentally analyzed for mathematical model. Second, convex model is applied for the prediction of required additional damping ratios to reduce responses below a specified target level. Chevron-type viscoelastic dampers are manufactured and installed at the first and second inter-stories, which are optimum places for response reduction. Sine-sweep and white noise excitations, which are generated by the hybrid mass-type actuator, are applied to the full scale building without and with dampers for performance verification. The transfer function of the building with four dampers, two of them installed at each first and second inter-story, are found to be lower than that of the building with two dampers installed at the first inter-story

  • PDF

하악구치 교합면의 design 조각법에 관한 연구 (A study on the design wax up technique for mandibular molar occlusion surface)

  • 문희경
    • 대한치과기공학회지
    • /
    • 제24권1호
    • /
    • pp.107-126
    • /
    • 2002
  • The first function of occlusion is mastication. Therefore the functional restoration of occlusal surface is very important. The restoration of occlusal surface is three method as wax bite technique, F.G.P. technique, cone technique. Many dental technician is using compound method. I am using compound method of wax bite technique and cone technique. I have knew common point on each teeth during I have waxing up wax pattern. So I studied on the design waxup technique for mandible molar occlusion. The results of the study were as follows; 1. The dam wax up method can restore axial contour of teeth very easy and make short working time of wax pattern. 2. The height of dam must be same with cusp of adjacent teeth. 3. Automatically the contour of tooth is appeared if the contour of dam is relationship with cuspid line of adjacent teeth. 4. The height of contour of buccal, lingual surface is formed natural curve to add fluid wax by gravitation. 5. The development groove of mandible first premolar is appeared V form. 6. The development groove of mandible second premolar is appeared Y form. 7, The development groove of mandible first molar is appeared M form. 8. The development groove of mandible second molar is W form. 9. The embrasure is formed to carve around contact point area as round convex. It affects to axial form of tooth. 10. The buccal, lingual groove of molar is formed parallel with direction of teeth arrangement.

  • PDF

단일 범용설비 운영을 위한 (r, Q) 정책 ((r, Q) Policy for Operation of a Multipurpose Facility)

  • 오근태
    • 한국경영과학회지
    • /
    • 제17권3호
    • /
    • pp.27-46
    • /
    • 1992
  • This paper considers an (r, Q) policy for operation of a multipurpose facility. It is assumed that whenever the inventory level falls below r, the model starts to produce the fixed amount of Q. The facility can be utilized for extra production during idle periods, that is, when the inventory level is still greater than r right after a main production operation is terminated or an extra production operation is finished. But, whenever the facility is in operation for an extra production, the operation can not be terminated for the main production even though the inventory level falls below r. In the model, the demand for the product is assumed to arrive according to a compound Poisson process and the processing time required to produce a product is assumed to follow an arbitary distribution. Similarly, the orders for the extra production is assumed to accur in a Poisson process are the extra production processing time is assumed to follow an arbitrary distribution. It is further assumed that unsatisfied demands are backordered and the expected comulative amount of demands is less than that of production during each production period. Under a cost structure which includes a setup/ production cost, a linear holding cost, a linear backorder cost, a linear extra production lost sale cost, and a linear extra production profit, an expression for the expected cost per unit time for a given (r, Q) policy is obtained, and using a convex property of the cost function, a procedure to find the optimal (r, Q) policy is presented.

  • PDF

전기 차량의 민감도 상관관계 (Sensitivity Correlations of Electrical Vehicle)

  • 이정익
    • 한국생산제조학회지
    • /
    • 제18권4호
    • /
    • pp.337-347
    • /
    • 2009
  • Generally, finite element models used in structural analysis have some uncertainties of the geometric dimensions, applied loads and boundary conditions, as well as in material properties due to the manufacturability of aluminum intensive body. Therefore, it is very important to refine or update a finite element model by correlating it with dynamic and static tests. The structural optimization problems of automotive body are considered for mechanical structures with initial stiffness due to preloading and in operation condition or manufacturing. As the mean compliance and deflection under preloading are chosen as the objective function and constraints, their sensitivities must be derived. The optimization problem is iteratively solved by a sequential convex approximation method in the commercial software. The design variables are corrected by the strain energy scale factor in the element levels. This paper presents an updated method based on the sensitivities of structural responses and the residual error vectors between experimental and simulation models.

  • PDF

유전자 알고리즘에 의한 HFC의 최적 제어파라미터 추정 및 설계 (Estimation of Optimal Control Parameters and Design of Hybrid Fuzzy Controller by Means of Genetic Algorithms)

  • 이대근;오성권;장성환;김용수
    • 대한전기학회논문지:시스템및제어부문D
    • /
    • 제49권11호
    • /
    • pp.599-609
    • /
    • 2000
  • The new design methodology of a hybrid fuzzy controller by means of the genetic algorithms is presented. First, a hybrid fuzzy controller(HFC) related to the optimal estimation of control parameters is proposed. The control input for the system in the HFC combined PID controller with fuzzy controller is a convex combination of the FLC's output and PID's output by a fuzzy variable, namely, membership function of weighting coefficient. Second, an auto-tuning algorithms utilizing the simplified reasoning method and genetic algorithms is presented to automatically improve the performance of hybrid fuzzy controller. Especially, in order to auto-tune scaling factors and PID parameters of HFC using GA, three kinds of estimation modes such as basic, contraction, and expansion mode are effectively utilized. The proposed HFC is evaluated and discussed to show applicability and superiority with the and of three representative processes.

  • PDF

CONTACT FORCE MODEL FOR A BEAM WITH DISCRETELY SPACED GAP SUPPORTS AND ITS APPROXIMATED SOLUTION

  • Park, Nam-Gyu;Suh, Jung-Min;Jeon, Kyeong-Lak
    • Nuclear Engineering and Technology
    • /
    • 제43권5호
    • /
    • pp.447-458
    • /
    • 2011
  • This paper proposes an approximated contact force model to identify the nonlinear behavior of a fuel rod with gap supports; also, the numerical prediction of interfacial forces in the mechanical contact of fuel rods with gap supports is studied. The Newmark integration method requires the current status of the contact force, but the contact force is not given a priori. Taylor's expansion can be used to predict the unknown contact force; therefore, it should be guaranteed that the first derivative of the contact force is continuous. This work proposes a continuous and differentiable contact force model with the ability to estimate the current state of the contact force. An approximated convex and differentiable potential function for the contact force is described, and a variational formulation is also provided. A numerical example that considers the particularly stiff supports has been studied, and a fuel rod with hardening supports was also examined for a realistic simulation. An approximated proper solution can be obtained using the results, and abrupt changes from the contacting state to non-contacting state, or vice versa, can be relieved. It can also be seen that not only the external force but also the developed contact force affects the response.

GUI를 이용한 철근콘크리트 연속보의 최적설계프로그램 개발 (Development of Optimum Design Program for Reinforced Concrete Continuous Beam Using Graphic User Interface System)

  • 조홍동;이상근;박중열;한상훈
    • 한국전산구조공학회논문집
    • /
    • 제13권1호
    • /
    • pp.1-11
    • /
    • 2000
  • 본 논문에서는 철근콘크리트 연속보의 최적설계 문제를 다루었으며, 설계실무에서 사용하기 위한 GUI 시스템을 개발하였다. 최적설계 문제형성을 위해 목적함수로 콘크리트, 철근 그리고 거푸집의 경비함수로 취하였고, 설계변수는 보의 폭과 유효높이 그리고 철근비이다. 제약조건으로 강도, 사용성, 내구성 그리고 기하학적인 조건을 고려하였다. 이렇게 형성된 비선형 최적설계문제는 축차선형계획기법과 축차볼록계획기법을 사용하여 최적해를 구하였으며 그 효율성을 비교하였다. 또한 실제 설계실무에서 손쉽게 사용할 수 있도록 visual basic을 이용한 GUI 시스템을 개발하여 수치예를 통한 적용성을 보였다.

  • PDF

Probabilistic penalized principal component analysis

  • Park, Chongsun;Wang, Morgan C.;Mo, Eun Bi
    • Communications for Statistical Applications and Methods
    • /
    • 제24권2호
    • /
    • pp.143-154
    • /
    • 2017
  • A variable selection method based on probabilistic principal component analysis (PCA) using penalized likelihood method is proposed. The proposed method is a two-step variable reduction method. The first step is based on the probabilistic principal component idea to identify principle components. The penalty function is used to identify important variables in each component. We then build a model on the original data space instead of building on the rotated data space through latent variables (principal components) because the proposed method achieves the goal of dimension reduction through identifying important observed variables. Consequently, the proposed method is of more practical use. The proposed estimators perform as the oracle procedure and are root-n consistent with a proper choice of regularization parameters. The proposed method can be successfully applied to high-dimensional PCA problems with a relatively large portion of irrelevant variables included in the data set. It is straightforward to extend our likelihood method in handling problems with missing observations using EM algorithms. Further, it could be effectively applied in cases where some data vectors exhibit one or more missing values at random.

eUCP 11조의 잠재적 하자의 규명과 합리적 개정방안의 모색 (A Study on the Inherent Defects in the eUCP Article 11 & Implication for the Revision)

  • 김기선
    • 무역상무연구
    • /
    • 제35권
    • /
    • pp.41-69
    • /
    • 2007
  • This study analyzes some important implications for the forthcoming revision of eUCP through the methodology of expected utility maximization theory. The overall results are as follows. First, beneficiary with an initial wealth has a risk-averse utility in traditional letter of credit transaction, and he would be more risk-averse in eUCP transaction. Secondly, the beneficiary who has risk-averse utility will pay for the risk premium to reduce the risk of corruption of an electronic record by means of cost of loss reduction activities. Thirdly, the cost of loss reduction activities is represented by a convex cost function, Fourthly, a risk averse beneficiary pursues loss reducing activities to the point where the expected marginal product of loss reduction is less than its marginal cost. Fifthly, a more risk-averse eUCP beneficiary will always select a higher level of loss reduction as long as the effectiveness of loss reduction is certain. Sixthly, when the effectiveness of loss reduction is uncertain, the more risk-averse eUCP beneficiary does not necessarily choose a higher level of loss reducing activities. Finally, it would be more reasonable that eUCP Article 11 should protect eUCP beneficiary who pursues a higher level of loss reducing activities.

  • PDF