• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.40 no.9
• /
• pp.859-865
• /
• 1991

This paper presents a new shaoe optimal design method using Desing Sensitivity Analysis. Design Sensitivity, defined as the variation of the object function with respect to the design variables, is based on the total differentiation of the matrix equation obtained from discretization of governing equations with respect to design variables. Analysis of the various characteristics adn calculation of Design Sensitivity of optimization model are achieved by using finite or boundary element methods. The proposee algorithm is applied to the optimal shape design of high voltage electrode under specified conditions. It is shown, from the numerical results, that the algorithm is very useful for the optimal shape design of electric apparatus.

• Structural Engineering and Mechanics
• /
• v.51 no.5
• /
• pp.707-725
• /
• 2014

In order to improve the crane-hook's performance and service life, we formulate a multi-criteria shape design problem considering practical conditions. The structural weight, the displacement at specified points and the induced matrix norm of stiffness matrix are adopted as the evaluation items to be minimized. The heights and widths of cross-section are chosen as the design variables. The design variables are expressed in terms of shape functions based on the Gaussian function. For this multi-objective optimization problem with three items, we utilize a multi-objective evolutionary algorithm, that is, the multi-objective Particle Swarm Optimization (MOPSO). As a common feature of obtained solutions, the side views are tapered shapes similar to those of actual crane-hook designs. The evaluation item values of the obtained designs demonstrate importance of the present optimization as well as the feasibility of the proposed optimal design approach.

• Asia pacific journal of information systems
• /
• v.11 no.1
• /
• pp.61-73
• /
• 2001

Predicting a variable using other variables in a large data set is a very difficult task. It involves selecting variables to include in a model and determining the shape of the relationship between variables. Nonparametric regression such as smoothing splines and neural networks are widely-used methods for such a task. We propose an alternative method based on a genetic algorithm(GA) to solve this problem. We applied GA to regression splines, a nonparametric regression method, to estimate functional forms between variables. Using several simulated and real data, our technique is shown to outperform traditional nonparametric methods such as smoothing splines and neural networks.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.24 no.2
• /
• pp.244-250
• /
• 2015

We propose a simulation technique to estimate the reduction effect of the pressure/flow pulsation by analysis of the pulsation variables and notch shape of the valve plate of a swash-plate-type variable piston pump. First, using SimulationX$^{(R)}$, we perform a theoretical kinematic analysis according to the variable swash-plate angle and rotational velocity in order to design a single-piston pump. In designing the notch shape of the valve plate of the swash-plate-type variable piston pump as one of the pulsation variables, we investigate the effect of the pulsation by comparing two notch types (circular type and V type). Then, we extend our analysis to a nine-piston pump model. This paper not only confirms the effect of the pressure/flow pulsation according to pulsation variables but can also be applied to the development of a SimulationX$^{(R)}$-based simulation technique for notch-shape optimization for a swash-plate-type variable piston pump.

• Proceedings of the SAREK Conference
• /
• 2009.06a
• /
• pp.1203-1208
• /
• 2009

In this paper, the flow characteristics of the mixed-flow pump impellers and diffusers were numerically predicted by commercial CFD software and DOE(design of experiments). We also discussed how to improve the performance of the mixed-flow pump by designing the impeller and diffuser in the mixed-flow pump. Geometric design variables were defined by the vane plane development which indicates the blade-angle distributions and length of the impeller and the diffusers. Firstly, the design optimization of the defined impeller geometric variables has been done. After that, the flow characteristics were analyzed in the point of incidence angle at the diffuser leading edge for the optimized impeller. Then design of the defined diffuser shape variables has been performed. The reason for the performance improvement was discussed by examining the flow characteristics through the diffuser.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.15 no.7 s.100
• /
• pp.775-780
• /
• 2005

This study uses a characteristic function to explain correlations between the objective function and design variables. Analysis of means and table of orthogonal array were carried out. The change of shape of diaphragm, thickness of diaphragm and voice coil weight based on the table of orthogonal array Is made. Therefore this study carried to decide design variables for minimizing 1st natural frequency and maximizing 2nd natural frequency of diaphragm using design of experiments and characteristic function with constraints. we showed improved design variables.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.22 no.3
• /
• pp.186-193
• /
• 2014

On-center handling feel indices are the important numerical factors that meaning performance of manuability and stability felt by drivers. They should be considered to suspension design. The major factor that affect to these indices is the shape of assist torque curve. This paper presents the variables that meaning the shape of the assist torque curve and analyze the effect of each variables. And using interaction formulas, it is proposed that the method to extract the variables that satisfy the goal indices. The result of simulation that applied assist torque curve extracted by the method demonstrate the capability of this method with error under 8%.

• Transactions of Materials Processing
• /
• v.19 no.5
• /
• pp.305-310
• /
• 2010

Recently industry of motor vehicle is making a gradual progress of automotive electric components. According to this step, wire harness equipped at motor vehicle is also increased. The most important component at the wire harness is electric connector. At the manufacturing process of electric connector, exactly at the crimping process, design variables, such as clamping-height, clamping-width and clamping die shape are critical parameters to assure satisfactory harness shape in clamping process of electric connector. In this study we have performed FEM simulation for clamping process and clarified the effect of design variables on compression rate of wire.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.2 no.4
• /
• pp.59-71
• /
• 1994

A multi-polynomial method is proposed to synthesize DRRD cam profiles. A cam lift duration s divided into 10 sections, each of them is expressed by a polynomial equation. 12 design variables are extracted from the cam profile displacement, velocity, and acceleration curves. Because all the design variables have physical meanings which are familiar to most cam designers, it is easy to imagine a profile shape from the design variables. The design envelope of the method is wide enough to be used in DRRD automotive cam designs. Polydyne cams, widely used in automotive engines, are included into the envelope. Unlike Polydyne cams, the method provides capability of wide velocity factor variations, which gives much flexibility in flat-faced tappet design. Area factor of profiles designed by the method can be increased 5-10% compared to those of Polydyne cams without increasing acceleration factor. The method is especially useful for cam profile optimizations.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.16 no.2
• /
• pp.169-180
• /
• 1992

The purpose of this study is to investigate the effects of type of clothing microclimate and exercise condition on ventilation. The experimental system employed a trace gas technique of the previous research. Clothing microclimate volume measurement was based on the substitution water technique for inter-clothing air volume. The experimental variables were tested at four levels of clothing microclimate spacing, microclimate shape of the previous research and two levels of exercise conditions. 2, 4, 6 cm ease were added to B/2+4 of basic blouse pattern for the microclimate spacing variable. Each combination of three variables were tested in triplicate. Analysis of variance of experimental variables on vetilation, such as oxygen exchange rate, half time of first order model was conducted. Oxygen exchange rate and half time of first order model are affected by the shape of microclimate air and exercise condition.