• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2022년도 추계학술대회
• /
• pp.123-125
• /
• 2022

본 논문에서는 Yolo-pose를 이용하여 장단기 메모리(Long short-term Memory)에 적용하는 시스템을 소개한다. 영상데이터로부터 Yolo-pose를 이용하여 일상생활과 낙상으로 구분된 데이터를 추출하여 LSTM에 적용하여 학습시킨다. 학습은 오버피팅을 방지하기 위하여 8대2의 Validation을 진행하며 Confusion matrix로 나타낸다. Yolo-pose의 결과값은 sensitivity와 specificity 모두 100%를 기록하여 일상생활과 낙상을 잘 구분하는 것을 확인 하였다.

• 한국복합재료학회:학술대회논문집
• /
• 한국복합재료학회 2003년도 춘계학술발표대회 논문집
• /
• pp.117-120
• /
• 2003

Electro-micromechanical techniques were applied using four-probe method for carbon nanotube (CNT) or nanofiber (CNF)/epoxy composites with their content. Carbon black (CB) was used to compare with CNT and CNF. The fracture of carbon fiber was detected by nondestructive acoustic emission (AE) relating to electrical resistivity for double-matrix composites test. Sensing for fiber tension was performed by electro-pullout test under uniform cyclic strain. The sensitivity for fiber damage such as fiber fracture and fiber tension was the highest for CNT/epoxy composites, and in CB case they were the lowest compared with CNT and CNF. Reinforcing effect of CNT obtained from apparent modulus measurement was the highest in the same content. The results obtained from sensing fiber damage were correlated with the morphological observation of nano-scale structure using FE-SEM. The information on fiber damage and matrix deformation and reinforcing effect of carbon nanocomposites could be obtained from electrical resistivity measurement as a new concept of nondestructive evaluation.

• Structural Engineering and Mechanics
• /
• 제22권4호
• /
• pp.401-418
• /
• 2006

Solutions to the problems of structural parameter estimation from modal response using leastsquares minimization of force or displacement residuals are generally sensitive to noise in the response measurements. The sensitivity of the parameter estimates is governed by the physical characteristics of the structure and certain features of the noisy measurements. It has been shown that the regularization method can be used to reduce effects of the measurement noise on the estimation error through adding a regularization function to the parameter estimation objective function. In this paper, we adopt the regularization function as the Euclidean norm of the difference between the values of the currently estimated parameters and the a priori parameter estimates. The effect of the regularization function on the outcome of parameter estimation is determined by a regularization factor. Based on a singular value decomposition of the sensitivity matrix of the structural response, it is shown that the optimal regularization factor is obtained by using the maximum singular value of the sensitivity matrix. This selection exhibits the condition where the effect of the a priori estimates on the solutions to the parameter estimation problem is minimal. The performance of the proposed algorithm is investigated in comparison with certain algorithms selected from the literature by using a numerical example.

• 대한기계학회논문집B
• /
• 제29권1호
• /
• pp.87-94
• /
• 2005

A computer code for Electrical Capacitance Tomography (ECT) is developed to sense the cross sectional phase distribution of two-phase flow in the rectangular pipe in which the tomography sensor furnished by the insulated wall, electrodes, and electric field screen. The computer code had two steps for the image reconstruction. In the forward projection step, the sensitivity matrix was constructed based on the electric field calculated by the finite difference method. In the backward projection step, the sensitivity matrix and the measured capacitances were used to reconstruct the cross sectional image. Several algorithms including LBP, TR, ITR, and PLI were employed to find the proper one for the two-phase flow analysis. Since the dielectric constant of the water in two-phase flow is sensitive to the thermal parameter such as, temperature and pressure, the developed code was evaluated to find their accuracy, speed of calculation, and sensitivity to the variation of the dielectric constant. It was found that the iterative methods are superior to the direct methods for the image reconstruction, and the PLI method was the best in the variation of the dielectric constants.

• 한국해양공학회지
• /
• 제26권5호
• /
• pp.11-17
• /
• 2012

In this study, a damage detection method that uses sensitivity-based finite (FE) element model updating with the natural frequency and zero frequency was proposed. The stiffness matrix for a structure was modified using the sensitivity-based FE model updating method. A sensitivity analysis was used to update the FE model, and the natural frequencies and zero frequencies were considered as target parameters to supplement the information on the vibration characteristics. The locations and values of the damages were estimated from the modified stiffness matrix. Several numerical examples were considered to verify the performance of the proposed method.

• Smart Structures and Systems
• /
• 제25권2호
• /
• pp.169-181
• /
• 2020

Sensor placement is a crucial aspect of bridge health monitoring (BHM) dedicated to accurately estimate and locate structural damages. In addressing this goal, a sensor placement framework based on the deflection influence line (DIL) analysis is here proposed, for the optimal design of damage detection-oriented BHM system. In order to improve damage detection accuracy, we explore the change of global stiffness matrix, damage coefficient matrix and DIL vector caused by structural damage, and thus develop a novel sensor placement framework based on the Fisher information matrix. Our approach seeks to determine the contribution of each sensing node to damage detection, and adopts a distance correction coefficient to eliminate the information redundancy among sensors. The proposed damage detection-oriented optimal sensor placement (OSP) method is verified by two examples: (1) a numerically simulated three-span continuous beam, and (2) the Pinghu bridge which has existing real damage conditions. These two examples verify the performance of the distance corrected damage sensitivity of influence line (DSIL) method in significantly higher contribution to damage detection and lower information redundancy, and demonstrate the proposed OSP framework can be potentially employed in BHM practices.

• 대한기계학회논문집A
• /
• 제23권11호
• /
• pp.1912-1921
• /
• 1999

The equipments mounted on guided-missile undertake heavy vibrational disturbance. Sometimes the equipments mounted on guided-missile go wrong so that the guided-missile flies over unintended place. For the vibration isolation of the equipments mounted on guided-missile, active vibration control was performed. In the case of active vibration technique, the stiffness matrix and the mass matrix are derived based on FEM (ANSYS5.0). Model reduction was carried out and, as a result, we got 7 DOF mass and stiffness matrix. For the sake of FEM model identification, modal experiment was carried out. With the help of Sensitivity Analysis, the natural frequencies of FEM were tuned to those of Experiment. In this work, the Sky Hook and the LQG control theory were adopted for v iteration control using stacked piezoactuator. Experiments were performed with changing excitation frequency from 10 Hz upto 200 Hz and we got frequency response function of guided-missile equipments. The magnitude of 3rd mode of guided-missile equipments is 8.6 % that of Uncontrolled in Skyhook controller and is 3.4 % that of uncontrolled in LQG controller.

• 대한전기학회논문지
• /
• 제39권8호
• /
• pp.792-804
• /
• 1990

This paper describes an efficient optimization algorithm by calculating sensitivity function for power system stabilization. In power system, the dynamic performance of exciter, governor etc. following a disturbance can be presented by a nonlinear differential equation. Since a nonlinear equation can be linearized for small disturbances, the state equation is expressed by a system matrix with system parameters. The objective function for power system operation will be related to the system parameter and the initial state at the optimal control condition for control or stabilization. The object function sensitivity to the system parameter can be considered to be effective in selecting the optimal parameter of the system.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2005년도 춘계 학술발표회 논문집
• /
• pp.327-334
• /
• 2005

In this paper, using an adjoint variable method, we develop a design sensitivity analysis (DSA) method applicable to 3-Dimensional heat conduction problems in steady state. Also, a topology design optimization method is developed using the developed DSA method. Design sensitivity expressions with respect to the thermal conductivity are derived. Since the already factorized system matrix is utilized to obtain the adjoint solution, the cost for the sensitivity computation is trivial. For the topology design optimization, the design variables are parameterized into normalized bulk material densities. The objective function and constraint are the thermal compliance of structures and allowable material volume, respectively, Through several numerical examples, the developed DSA method is verified to yield efficiency and accurate sensitivity results compared with finite difference ones. Also, the topology optimization yields physical meaningful results.

• 동력기계공학회지
• /
• 제4권2호
• /
• pp.71-76
• /
• 2000

It is well known that finite element analysis often has the inaccuracy when they are in conflict with test results. Model updating is concerned with the correction of analytical model by processing records of response from test results. This paper introduce a model updating technique using the frequency response function data. The measurement data is able to be used directly in the FRF sensitivity method because it is not necessary to identify. When a damping model is updated, it is necessary for the sensitivity matrix to be divided Into the complex part and real part. As an applying model, a cantilever and a rotor system are used. Specially the machined clearance($C_p$) of the journal bearing is updated.