• 한국정밀공학회지
• /
• 제25권7호
• /
• pp.111-115
• /
• 2008

This paper describes the optimization of injection molding conditions to minimize sink marks. Sink marks, which refer to a small depression on the surface opposite a thick wall thickness, are often encounted in injection molded plastic parts. Part geometry, material properties and processing conditions during injection molding can affect the sink mark depth. We designed the runner system which is possible balanced filling to cavities using CAE program $Moldflow^{TM}$ and then obtained optimal processing conditions by Taguchi's Robust Design technique. By actual injection molding using optimized mold and molding conditions, it confirmed that sink mark depth decreased zero compared to 1mm level in the conventional mold and process.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2004년도 춘계학술대회
• /
• pp.733-738
• /
• 2004

This paper deals with feedback control of a hydraulic unit for direct yaw moment control, a method used to actively maintain the dynamic stability of an automobile. The uncertain parameters and complex structure naturally call for empirical modeling of the hydraulic unit, which readily results in a control-oriented model with high fidelity. The identified model is cross-validated against experimental data under various conditions, which helps to establish model uncertainty. Then, the $H_{\infty}$ optimization technique is employed to synthesize a controller with guaranteed robust stability and performance against the model uncertainty. The performance of the synthesized controller is verified using experimental results, which shows the viability of the proposed approach in a real-world application.

• 한국공간구조학회:학술대회논문집
• /
• 한국공간구조학회 2006년도 춘계 학술발표회 논문집 제3권1호(통권3호)
• /
• pp.227-232
• /
• 2006

This paper investigate the optimum thickness distribution of plate structure with different essential boundary conditions in the fundamental natural frequency maximization problem. In this study, the fundamental natural frequency is considered as the objective function to be maximized and the initial volume of structures is used as the constraint function. The computer-aided geometric design (CAGD) such as Coon's patch representation is used to represent the thickness distribution of plates. A reliable degenerated shell finite element is adopted calculate the accurate fundamental natural frequency of the plates. Robust optimization algorithms implemented in the optimizer DoT are adopted to search optimum thickness values during the optimization iteration. Finally, the optimum thickness distribution with respect to different boundary condition

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1996년도 하계학술대회 논문집 B
• /
• pp.1158-1160
• /
• 1996

In this paper, the optimization of fuzzy controller using genetic algorithm is studied. The fuzzy controller has been widely applied to industries because it is highly flexible, robust, easy to implement, and suitable for complex systems. Generally, the design of fuzzy controller has difficulties in determining the structure of the rules and the membership functions. To solve these problems, the proposed method optimizes the structure of fuzzy roles and the parameters of membership functions simultaneously in so off-line method. The proposed method is evaluated through computer simulations.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2013년도 추계학술대회 논문집
• /
• pp.23-24
• /
• 2013

In recent years, ISG (Integrated Starter and Generator) system receives a great attention for electric electrification of normal gasoline vehicle. As a cost-effect machine design, an ISG without a permanent magnet is considered. A 12slot-10pole field-excitation flux-switching synchronous machine (FEFSSM) is designed and analyzed via JMAG. The active parts such as the field excitation coil and armature coil are located on the stator. The rotor part consisting of single piece iron makes it more robust and suitable to apply for high speed motor drive system application coupled with reduction belt. The design target is the motor with a maximum torque of 40Nm, a maximum power of 10kW and a maximum speed of 14000 rpm. In this paper, design optimization method is proposed for high torque capability.

• Structural Engineering and Mechanics
• /
• 제14권2호
• /
• pp.223-236
• /
• 2002

The main objective of the present paper is to address a formal procedure for orthotropic steel plates design. The theme of the proposed approach is to recast the design procedure into a mathematical programming model. The objective function to be optimized is the total weight of the structure. The total weight is function of its layout parameters and structural element design variables. Mean while the proposed approach takes into consideration the strength and rigidity criteria in addition to other dimensional constraints. A nonlinear programming model is developed which consists of a nonlinear objective function and a set of implicit/explicit nonlinear constraints. A transformation method is adopted for minimization strategy, where the primal model constrained problem is transformed into a sequence of unconstrained minimization models. The search strategy is based on the well-known Fletcher/Powell algorithm. The finite element technique is adopted for discretization and analysis strategies. Mindlin theory is selected to simulate the finite element model and a selective reduced integration scheme is exploited to avoid a shear lock problem.

• 산업경영시스템학회지
• /
• 제24권63호
• /
• pp.1-12
• /
• 2001

When designing the parameter on the multiple quality characteristics, there has been a study for optimization of problems, but there has been few former study on the possible conflicting phenomena in consideration of the correlations among the characteristics. To solve the issue on the optimal design for multiple quality characteristics, this study propose the expected loss function with cross-product terms among the characteristics and derived range of the coefficients of the terms. The model will be used to determine the global optimal design parameters where there exists the conflict among the characteristics, which shows difference in optimal design parameters for the individual characteristics. Also, when an adjustment factor exists for each m performance characteristics, this research propose a method by minimizing the expected loss after adjustment.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.7-9
• /
• 2004

This paper provides an observer-based $H_{\infty}$ controller design method for singular systems with and without time-varying delay by just one LMI condition. The sufficient condition for the existence of controller and the controller design method are presented by perfect LMI (linear matrix inequality) approach. The design procedure involves solving an LMI. The observer-based $H_{\infty}$ controller in the existing results can be constructed from the coupled two or more conditions while the proposed controller design method can be obtained from an LMI condition, which can be solved efficiently by convex optimization. Since the obtained condition can be expressed as an LMI form, all variables including feedback gain and observer gain can be calculated simultaneously by Schur complement and changes of variables. An example is given to illustrate the results.

• 한국기계가공학회지
• /
• 제20권11호
• /
• pp.108-114
• /
• 2021

In this study, a multivariate function was applied to the genetic algorithm for D-type lugs currently used in shipyards to closely analyze the behavioral form of weight loss without double plates. An optimal lifting lug structure design without attachments is proposed. MATLAB R2016a was used to design features by applying multivariate functions to genetic algorithms. Furthermore, the design was achieved by deriving the optimal shapes of lugs using genetic algorithms. The shapes of the designed lugs were validated for structural bonding using the structural analysis program ANSYS 2020 R2, and a robust design of lugs with no appendages was developed.

• Journal of Electrical Engineering and Technology
• /
• 제12권6호
• /
• pp.2388-2398
• /
• 2017

In this study, we propose a robust face recognition system to pose variations based on automatic pose estimation. Radial basis function neural network is applied as one of the functional components of the overall face recognition system. The proposed system consists of preprocessing and recognition modules to provide a solution to pose variation and high-dimensional pattern recognition problems. In the preprocessing part, principal component analysis (PCA) and 2-dimensional 2-directional PCA ($(2D)^2$ PCA) are applied. These functional modules are useful in reducing dimensionality of the feature space. The proposed RBFNNs architecture consists of three functional modules such as condition, conclusion and inference phase realized in terms of fuzzy "if-then" rules. In the condition phase of fuzzy rules, the input space is partitioned with the use of fuzzy clustering realized by the Fuzzy C-Means (FCM) algorithm. In conclusion phase of rules, the connections (weights) are realized through four types of polynomials such as constant, linear, quadratic and modified quadratic. The coefficients of the RBFNNs model are obtained by fuzzy inference method constituting the inference phase of fuzzy rules. The essential design parameters (such as the number of nodes, and fuzzification coefficient) of the networks are optimized with the aid of Particle Swarm Optimization (PSO). Experimental results completed on standard face database -Honda/UCSD, Cambridge Head pose, and IC&CI databases demonstrate the effectiveness and efficiency of face recognition system compared with other studies.