• 수산경영론집
• /
• 제44권3호
• /
• pp.77-84
• /
• 2013

While some steps in laver aquaculture production can be controlled artificially to a certain extent, the culturing process is largely affected by natural factors, such as the characteristics of seawater, climatic and oceanographic conditions, etc. This study aims to find a direct relationship between climatic and oceanographic factors (water temperature, air temperature, salinity, rainfall, sunshine duration and wind speed) and laver aquaculture production in Wando region, the biggest aquaculture production area of laver, located in the southwest coast of Korea using a multiple regression analysis. Despite the small sample size of a dependent variable, the goodness of model fit appeared acceptable. In addition, the R-squared value was 0.951, which means that the variables were very explanatory. Model results indicated that duration of sunshine, temperature, and rainfall during the farming period from the end of September to the end of April would be important factors affecting significantly to the laver aquaculture production.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2002년도 추계학술대회논문집
• /
• pp.605-610
• /
• 2002

Eigenvalues are taken as performance criteria for structural damping design using viscoelastic material. Given material properties, optimal distribution of damping material is sought based on eigenvalue sensitivity. For eigenanalysis of frequency dependent viscoelastic material treated structures, Golla-Mushes-McTavish (GHM) model is used and some dominant modes are chosen for consideration. To avoid the intensity of computation caused by increased problem size, an alternative approximate method is proposed which uses elastic modes and can be applied under small damping assumption. A cantilever beam treated with unconstrained viscoelastic layer is tested and optimal distribution of thickness of the layer is illustrated. Partial coverage configurations are compared with the one-sided full coverage case.

• 한국자동차공학회논문집
• /
• 제16권4호
• /
• pp.29-37
• /
• 2008

Vehicle stability is a very important subject in vehicle design and control, because vehicle safety is closely dependent upon its dynamic stability. For the vehicle stability analysis, the nonlinear vehicle model of a mid-size car with three DOF - longitudinal, lateral and yaw - is employed. A rigorous method is used to determine the vehicle stability region in plane motion. An algorithm is used to materialize a topology theorem, which enables to find the exact stability region. A stability criterion for the critical cornering is proposed.

• 한국경영과학회지
• /
• 제14권1호
• /
• pp.88-96
• /
• 1989

General measure in the reliability is the k-terminal reliability, which is the probability that the specified vertices are connected by the working edges. To compute the k-terminal reliability components are usually assumed to be statistically independent. In this study the modeling and analysis of the k-terminal reliability are investigated when dependency among components is considered. As the size of the network increases, the number of the joint probability parameter to represent the dependency among components is increasing exponentially. To avoid such a difficulty the structured-event-based-reliability model (SERM) is presented. This model uses the combination of the network topology (physical representation) and reliability block diagram (logical representation). This enables us to represent the dependency among components in a network form. Computational algorithms for the k-terminal reliability in SERM are based on the factoring algorithm Two features of the ractoring algorithm are the reliability preserving reduction and the privoting edge selection strategy. The pivoting edge selction strategy is modified by two different ways to tackle the replicated edges occuring in SERM. Two algorithms are presented according to each modified pivoting strategy and illustrated by numerical example.

• 제어로봇시스템학회논문지
• /
• 제18권6호
• /
• pp.519-525
• /
• 2012

Electric brake booster systems replace conventional pneumatic brake boosters with electric motors and rotary-todisplacement mechanisms including ECU (Electronic Control Unit). Electric booster brake systems require precise target pressure tracking and control robustness because vehicle brake systems operate properly given the large range of loading and temperature, actuator saturation, load-dependent friction. Also for the implement of imbedded control system, the controller should be selected considering the limited memory size and the cycle time problem of real brake ECU. In this study, based on these requirements, a sliding mode controller has been chosen and applied considering both model uncertainty and external disturbance. A mathematical model for the electric booster is derived and simulated. The developed sliding mode controller considering chattering problem has been compared with a conventional cascade PID controller. The effectiveness of the controller is demonstrated in some braking cases.

• 대한기계학회논문집B
• /
• 제29권2호
• /
• pp.287-293
• /
• 2005

In the present study, to determine the flow rate of droplets supplied to heat transfer surface after (j-1)th rebound, $D_X[j{\ge}2]^{\ast}$, it was assumed that the rebound droplets are distributed according to the Gaussian distribution from 0 to L, in which the flight distance L is determined by maximum flight distance $L_{max}$. We also assumed that $L_{max}$ is dependent on the air flow velocity and mean size of droplets. The local heat flux of a dilute spray in high temperature region was predicted using the newly evaluated $D_X[j{\ge}2]^{\ast}$. In addition, the predicted results by the present model were compared with the existing experimental data.

• Structural Engineering and Mechanics
• /
• 제63권2호
• /
• pp.137-146
• /
• 2017

The size-dependent behavior of single walled carbon nanotubes (SWCNT) embedded in the elastic medium and subjected to the initial axial force is investigated using the mixed finite element method. The SWCNT is assumed to be Euler-Bernoulli beam incorporating nonlocal theory developed by Eringen. The mixed finite element model shows its great advantage of dealing with nonlocal behavior of SWCNT subjected to a concentrated load owing to the existence of two coefficients ${\alpha}_1$ and ${\alpha}_2$. This is the first numerical approach to deal with a puzzling fact of nonlocal theory with concentrated load. Numerical examples are performed to show the accuracy and efficiency of the present method. In addition, parametric study is carefully carried out to point out the influences of nonlocal effect, the elastic medium, and the initial axial force on the behavior of the carbon nanotubes.

• 한국멀티미디어학회논문지
• /
• 제16권9호
• /
• pp.1018-1030
• /
• 2013

The corpus callosum is the largest connective structure in the brain, and its shape and size are correlated to sex, age, brain growth and degeneration, handedness, musical ability, and neurological diseases. Manually segmenting the corpus callosum from brain magnetic resonance (MR) image is time consuming, error prone, and operator dependent. In this paper, two semi-automatic segmentation methods are present: the active contour model-based approach and the active shape model-based approach. We tested these methods on an MR image of the human brain and found that the active contour approach had better segmentation accuracy but was slower than the active shape approach.

• 대한기계학회논문집
• /
• 제18권12호
• /
• pp.3202-3218
• /
• 1994

Mode shapes and natural frequencies of vibrating laminated composite plates are taken using real-time and time-average holographic interferometry. Debonds and delamination in the laminated plates are measured nondestructively. During holographic testing of composite plates, it has been found that the conditions for the local resonance in debonds are strongly dependent on the frequency of excitation. A membrane resonance model was proposed to describe this behavior. Relations between characteristic length according to the size, shape of debonds and membrane resonance frequency are presented. Several experiments were performed to verify the membrane resonance model. The agreements between the predicted excitation frequency and the observed resonance frequency are good. The experimental results show that higher stresses and strains due to local resonance lead to the debond detection.

• 전자공학회논문지B
• /
• 제32B권11호
• /
• pp.1489-1495
• /
• 1995

In this paper, we propose recurrent neural prediction models (RNPM), recurrent neural networks trained as a nonlinear predictor of speech, as a new connectionist model for speech recognition. RNPM modulates its mapping effectively by internal representation, and it requires no time alignment algorithm. Therefore, computational load at the recognition stage is reduced substantially compared with the well known predictive neural networks (PNN), and the size of the required memory is much smaller. And, RNPM does not suffer from the problem of deciding the time varying target function. In the speaker dependent and independent speech recognition experiments under the various conditions, the proposed model was comparable in recognition performance to the PNN, while retaining the above merits that PNN doesn't have.