• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1994.04a
• /
• pp.410-420
• /
• 1994

In many project-oriented production systems, e.g., shipyards or large-scale steel products manufacturing, resource loading by an activity is flexible, and the activity duration is a function of resource allocation. For example, if one doubles the size of the crew assigned to perform an activity, it may be feasible to complete the activity in half the duration. Such flexibility has been modeled by Weglarz [131 and by Leachman, Dincerler, and Kim [7] in extended formulations of the resource-constrained project scheduling problem. This paper presents a new algorithmic approach to the problem that combines the ideas proposed by the aforementioned authors. The method we propose involves a two-step approach: (1) solve the resource-constrained scheduling problem using a heuristic, and (2) using this schedule as an initial feasible solution, find improved resource allocations by solving a linear programming model. We provide computational results indicating the superiority of this approach to previous methodology for the resource-constrained scheduling problem. Extensions to the model to admit overlap relationships of the activities also are presented.

• International journal of advanced smart convergence
• /
• v.2 no.1
• /
• pp.18-20
• /
• 2013

The significance, is to introduce a novel way to employ the improved Runge-Kutta fifth order five stage method, here after called as Modified IRK(5,5) method, for system of second order robot arm problem and variations in angles at the joints in which parameters governing with two degrees of freedom which requires lesser number of function evaluations per time step as compared to the existing ones, in order to save time and spaceAn ultimate aim of this present paper is to solve application problem such as robot arm and initial value problems by applying Runge-Kutta fifth order five stage numerical techniques. The calculated output for robot arm coincides with exact solution which is found to be better, suitable and feasible for solving real time problems.

• Journal of the Korean Society of Safety
• /
• v.6 no.4
• /
• pp.34-44
• /
• 1991

In this study, the general solution of heat balance equation including conductive heat loss were suggested and were determined the constants with the results of experiment in hot tunnel in order to derive the general equation for the response time and to investigate the response time index which represent the characteristics of response of sprinkler head in actual fires. Two types of test were considered, the plunge test, in which the air temperature is represented by a step function, and the ramp test, in which the air temperature increases at a constant rate. As a result, simple equations were derived, which can be predicted the response time for the ramp type fire with the rate of temperature rise and gas velocity, for the plunge type fire with temperature and gas velocity. Also other useful data, such as the effective temperature, time constant, response time index and conduction parameter were obtained.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.2
• /
• pp.336-347
• /
• 1992

The three dimensional inviscid transonic cascade flow was investigated numerically, incorporation a four stage Runge-Kutta integration method proposed by Jameson. Time marching to the steady state was accelerated by using optimum time step and enthalpy damping. In describing the boundary conditions at inlet and outlet, Riemann invariants are considered. By adding a second and a fourth order artificial viscocities, the numerical instability due to the propagation of undamped disturbance or the rapid change of state near the shock has been prevented. The numerical results for are bump cascade, cambered two dimensional turbine cascade and three dimensional stator cascade agreed reasonably well with previous results. It has been known that the accuracy of the solution depended a lot on the modeling of the leading or trailing edge.

• Management Science and Financial Engineering
• /
• v.16 no.1
• /
• pp.81-117
• /
• 2010

Semi-supervised learning incorporates unlabeled examples, whose labels are unknown, as well as labeled examples into learning process. Although transductive support vector machine (TSVM), one of semi-supervised learning models, was proposed about a decade ago, its application to large-scaled data has still been limited due to its high computational complexity. Our previous research addressed this limitation by introducing a branch-and-bound algorithm for finding an optimal solution to TSVM. In this paper, we propose three new techniques to enhance the performance of the branch-and-bound algorithm. The first one tightens min-cut bound, one of two bounding strategies. Another technique exploits a graph-based approximation to a support vector machine problem to avoid the most time-consuming step. The last one tries to fix the labels of unlabeled examples whose labels can be obviously predicted based on labeled examples. Experimental results are presented which demonstrate that the proposed techniques can reduce drastically the number of subproblems and eventually computational time.

• International Journal of Precision Engineering and Manufacturing
• /
• v.9 no.2
• /
• pp.44-49
• /
• 2008

This paper describes the problem of ultraprecision positioning with a ball screw mechanism in the microdynamic range, along with its solution. We compared the characteristics of two ball screw mechanisms with different table masses. The experimental results showed that the vibration resulting from the low stiffness of the ball screw degraded the positioning performance in the microdynamic range for the heavyweight mechanism. The proposed nominal characteristic trajectory following (NCTF) controller was designed for ultra precision positioning of the ball screw mechanism. The basic NCTF control system achieved ultra precision positioning performance with the lightweight mechanism, but not with the heavyweight mechanism. A conditional notch filter was added to the NCTF controller to overcome this problem. Despite the differences in payload and friction, both mechanisms then showed similar positioning performance, demonstrating the high robustness and effectiveness of the improved NCTF controller with the conditional notch filter. The experimental results demonstrated that the improved NCTF control system with the conditional notch filter achieved ultra precision positioning with a positioning accuracy of better than 10 nm, independent of the reference step input height.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.578-578
• /
• 2012

Graphene, a flat one-atom-thick two-dimensional layer of carbon atoms, is considered to be a promising candidate for nanoelectronics due to its exceptional electronic properties. Most of all, future nanoelectronics such as flexible displays and artificial electronic skins require low cost manufacturing process on flexible substrate to be integrated with high resolutions on large area. The solution based printing process can be applicable on plastic substrate at low temperature and also adequate for fabrication of electronics on large-area. The combination of printed electronics and graphene has allowed for the development of a variety of flexible electronic devices. As the first step of the study, we prepared the gate electrodes by printing onto the gate dielectric layer on PET substrate. We showed the performance of graphene field-effect transistor with electrohydrodynamic (EHD) inkjet-printed Ag gate electrodes.

• Journal of the Korean Ceramic Society
• /
• v.32 no.7
• /
• pp.858-864
• /
• 1995

A porous glass-ceramics body was prepared in the phosphate system. The glass composition of 47.2CaO-22.2TiO2-30.6P2O5 (mol%) containing a few weight percent of ZrO2 was suitable for a mother glass of a porous glass-ceramics. The dense glass-ceramics body was made by a two-step heat treatment of the mother glass. The crystalline phases of the glass-ceramics were $\beta$-Ca3(PO4)2 and CaTi4(PO4)6. The $\beta$-Ca3(PO4)2 phase could be selectively leached out with HCl solution and thus a crystalline $\beta$-Ca3(PO4)2 skeleton was remained. The dimension and shape of the porous glass-ceramics were nearly the same as the those of the first formed glass. The specific surface area and average proe radius of the porous glass-ceramics were 19$m^2$/g and 22 nm, respectively.

• Nuclear Engineering and Technology
• /
• v.32 no.1
• /
• pp.46-56
• /
• 2000

Numerical results are presented for the 2-dimensional turbulent transient heat transfer of the shell/tube heat exchanger with a step change of the inlet temperature in the primary side. Heat transfer boundary conditions outside the pipe are given partially by the convection heat transfer conditions and partially by insulated conditions. Calculation results were obtained by solving the unsteady two-dimensional elliptic forms for the Reynolds-averaged governing equations for the mass, momentum and energy. Finite-difference method was used to obtain discretization equations, and the SIMPLER solution algorithm was employed for the calculation procedure. Turbulent model used is the algebraic model proposed by Cebeci-Smith. Results presented include the time variant Nusselt number distribution, average temperature distribution and outlet temperatures for the various inlet temperatures and flow rates.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2009.05a
• /
• pp.432-434
• /
• 2009

Metal nano-wire arrays on Cu-coated seed layers were fabricated by aqueous solution method using sulfate bath at room temperature. The seed layers were coated on Anodic aluminum oxide (AAO) bottom substrates by electrochemical deposition technique, length and diameter of metal nano-wires were dominated by controlling the deposition parameters, such as deposition potential and time, electrolyte temperature. Anodic aluminum oxide (AAO) was used as a template to prepare highly ordered Ni, Fe, Co and Cu multilayer magnetic nano-wire arrays. This template was fabricated with two-step anodizing method, using dissimilar solutions for Al anodizing. The pore of anodic aluminum oxide templates were perfectly hexagonal arranged pore domains. The ordered Ni, Fe, Co and Cu systems nano-wire arrays were characterized by Field Emission Scanning Electron Microscopy (FE-SEM) and Vibrating Sample Magnetometer (VSM). The ordered Ni, Fe, Co and Cu systems nano-wires had different preferred orientation. In addition, these nano-wires showed different magnetization properties under the electrodepositing conditions.