• 한국해양공학회지
• /
• 제26권6호
• /
• pp.80-85
• /
• 2012

The methods of robust design optimization (RDO) and reliability-based robust design optimization (RBRDO) were implemented in the present study. RBRDO is an integrated method that accounts for the design robustness of an objective function and for the reliability of constraints. The objective function in RBRDO is expressed in terms of the mean and standard deviation of an original objective function. Thus, a multi-objective formulation is employed. The regressive approximate models are generated via the moving least squares method (MLSM) and constraint-feasible moving least squares method (CF-MLSM), which make it possible to realize the feasibility regardless of the multimodality/nonlinearity of the constraint function during the approximate optimization processes. The regression model based RBRDO is newly devised and its numerical characteristics are explored using the design of an actively controlled ten bar truss structure.

• 한국도로학회논문집
• /
• 제15권6호
• /
• pp.125-133
• /
• 2013

PURPOSES : The purpose of this study is to propose delay-minimizing operation methodology of a signalized intersection based upon optimization of lane-uses on approaching lanes for an intersection. METHODS : For the optimization model of lane-uses, a set of constraints are set up to ensure feasibility and safety of the lane-uses, traffic flow, and signal settings. Minimization of demand to saturation flow ratio of a dual-ring signal control system is introduced to the objective function for delay minimization and effective signal operation. Using the optimized lane-uses, signal timings are optimized by delay-based model of TRANSYT-7F. RESULTS : It was found that the proposed objective function is great relation with delay time for an intersection. From the experimental results, the method was approved to be effective in reducing delay time. Especially, cases for two left-turn lanes reduced greater delays than those for a left turn lane. It is noticed that the cases for different traffic volume by approach reduced greater delays than those for the same traffic volume by approach. CONCLUSIONS : It was concluded that the objective function is proper for lane-uses optimizing model and the operation method is effective in reducing delay time for signalized intersections.

• 대한조선학회논문집
• /
• 제42권2호
• /
• pp.175-182
• /
• 2005

It is necessary to develop an efficient optimization technique to optimize the engineering structures which have given design spaces, discrete design values and several design goals. As optimization techniques, direct search method and stochastic search method are widely used in designing of engineering structures. The merit of the direct search method is to search the optimum points rapidly by considering the search direction, step size and convergence limit. And the merit of the stochastic search method is to obtain the global optimum points by spreading point randomly entire the design spaces. In this paper, a Pareto optimal based multi-objective function method (PMOFM) is developed by considering the search direction based on Pareto optimal points, step size, convergence limit and random search generation . The PMOFM can also apply to the single objective function problems, and can consider the discrete design variables such as discrete plate thickness and discrete stiffener spaces. The design results are compared with existing Evolutionary Strategies (ES) method by performing the design of double bottom structures which have discrete plate thickness and discrete stiffener spaces.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.1927-1932
• /
• 2003

In a previous study, the authors have proposed the Cooperative Collision Avoidance (CCA) method which enables mobile robots to cooperatively avoid collisions, by extending the concept of the Velocity Obstacle to multiple robot systems. The method introduced an evaluation function considering an operation objective so that each robot can choose the velocity which optimizes the function. As the evaluation function could be of an arbitrary type, this method is applicable to a wide variety of tasks. However, it complicates the optimization of the function especially in real-time. In addition, construction of the evaluation function requires an operation objective of the other robot which is very hard to obtain without communication. In this paper, the CCA method is improved considering such problems for implementation. To decrease computational costs, the previous method is simplified by introducing two essential assumptions. Then, by treating the desired direction of locomotion for each robot as the operation objective, an operation objective estimator which estimates the desired direction of the other robot is introduced. The only measurement required is the other robot's relative position, since the other information can be obtained through the estimation. Hence, communicational devices that are necessary for most other cooperative methods are not required. Moreover, mobile robots employing the method can avoid collisions with uncooperative robots or moving obstacles as well as with cooperative robots. Consequently, this improved method can be applied to general dynamic environments consisting of various mobile robots.

• 한국수자원학회논문집
• /
• 제38권10호
• /
• pp.861-869
• /
• 2005

강우-유출모형은 적용대상 유역이 가지고 있는 수문학적 성질을 최대한 반영할 수 있도록 보정되어야 한다. 본 연구에서는 SSARR 모형의 5개의 매개변수를 안동댐 상류유역에 보정하기 위하여 다중목적함수와 유전자알고리즘을 이용하였다. 다중목적함수의 최적해는 단일한 매개변수로 이루어지는 것이 아니라 다양한 목적함수들에 따라서 결정되는 파레토 최적해로 구성된다. 다중목적함수를 이용한 모형의 보정방법은 보정시간 및 작업 반복에 따른 노력을 감소시킬 수 있었으며, 파레토 최적해를 사용함으로써 적용 목적에 따라 최대유랑을 잘 모의할 필요가 있다거나 전체 체적을 잘 모의할 필요가 있는 경우에 적합한 매개변수를 사용자가 선택하여 사용할 수 있는 장점이 있다.

• 대한기계학회논문집A
• /
• 제37권5호
• /
• pp.589-599
• /
• 2013

강건최적설계의 개념은 다구찌 법에 근간을 두고 있다. 특히, 목적함수의 강건성 지수들은 설계변수나 파라미터의 변동에 둔감하고 보수적인 설계를 추구한다. 그 목적을 달성하기 위해 다양한 강건성 지수들이 소개되고 있다. 소개된 다양한 지수와 방법은 나름의 목적과 의미를 지니고 있다. 하지만, 다구찌 법에서 의미하는 강건설계의 의미를 목적함수의 강건성 지수로 반영하여 최적설계 문제로 확장하는 것에는 한계점이 발생할 수 있다. 본 논문의 목적은 기존 강건성 지수 연구들의 특징과 한계점을 파악하고 강건최적설계 연구의 고찰을 수행하는데 있다. 목적함수의 강건성 지수들의 특징을 확인하기 위해 결정론적 최적해와 강건해의 구분이 명확한 수학적 예제를 사용하여 평가를 수행하고 분석하였다. 더불어, 고찰을 토대로 강건최적설계에서의 강건성에 대한 새로운 관점과 상한함수를 사용한 목적함수의 강건성 지수를 제시하였다.

• Ocean Systems Engineering
• /
• 제9권2호
• /
• pp.111-133
• /
• 2019

Flow through a scaled horizontal axis marine current turbine was numerically simulated after validation and the turbine design was optimized. The computational fluid dynamics (CFD) code Ansys-CFX 16.1 for numerical modeling, an in-house blade element momentum (BEM) code for analytical modeling and an in-house surrogate-based optimization (SBO) code were used to find an optimal turbine design. The blade-pitch angle (${\theta}$) and the number of rotor blades (NR) were taken as design variables. A single objective optimization approach was utilized in the present work. The defined objective function was the turbine's power coefficient ($C_P$). A $3{\times}3$ full-factorial sampling technique was used to define the sample space. This sampling technique gave different turbine designs, which were further evaluated for the objective function by solving the Reynolds-Averaged Navier-Stokes equations (RANS). Finally, the SBO technique with search algorithm produced an optimal design. It is found that the optimal design has improved the objective function by 26.5%. This article presents the solution approach, analysis of the turbine flow field and the predictability of various surrogate based techniques.

• International Journal of Contents
• /
• 제10권1호
• /
• pp.23-28
• /
• 2014

When developing a classifier using various objective functions, it is important to compare the performances of the classifiers. Although there are statistical analyses of objective functions for classifiers, simulation results can provide us with direct comparison results and in this case, a comparison criterion is considerably critical. A Receiver Operating Characteristics (ROC) graph is a simulation technique for comparing classifiers and selecting a better one based on a performance. In this paper, we adopt the ROC graph to compare classifiers trained by mean-squared error, cross-entropy error, classification figure of merit, and the n-th order extension of cross-entropy error functions. After the training of feed-forward neural networks using the CEDAR database, the ROC graphs are plotted to help us identify which objective function is better.

• 한국공작기계학회논문집
• /
• 제16권4호
• /
• pp.36-44
• /
• 2007

This paper is focused on the optimum design for decreasing volume and increasing reliability of multi-stage gear drive. For the optimization on volume and reliability, multi-objective optimization is used. The genetic algorithm is introduced to multi-objective optimization method and it is used to develop the optimum design program using exterior penalty function method to solve the complicated subject conditions. A 5 staged gear drive(geared motor) is chosen to compare the result of developed optimum design method with the existing design. Each of the volume objective, reliability objective, and volume-reliability multi-objectives are performed and compared with existing design. As a result, optimum solutions are produced, which decrease volume and increase reliability. It is shown that the developed design method is good for multi-stage gear drive design.

• International Journal of Contents
• /
• 제8권4호
• /
• pp.30-35
• /
• 2012

Error surfaces provide us with very important information for training of feed-forward neural networks (FNNs). In this paper, we draw the contour plots of various error or objective functions for training of FNNs. Firstly, when applying FNNs to classifications, the weakness of mean-squared error is explained with the viewpoint of error contour plot. And the classification figure of merit, mean log-square error, cross-entropy error, and n-th order extension of cross-entropy error objective functions are considered for the contour plots. Also, the recently proposed target node method is explained with the viewpoint of contour plot. Based on the contour plots, we can explain characteristics of various error or objective functions when training of FNNs proceeds.