• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.19 no.1
• /
• pp.50-56
• /
• 2010

This paper presents a new design algorithm for piece-removing dynamical system, based on 6-Sigma DMADOV technique using ARIZ and Brainstorming. Our design target is the piece-removing system installed on a mobile platform of bead-grinding equipment. The 6-Sigma DMADOV technique guides us design process according to 6 steps, i.e., Define - Measure - Analyze - Design - Optimize - Verify. A Design strategy to reduce the weight of piece-removing dynamical system will be explored by using ARIZ, i.e.,(the abbreviation of Algorithm for Inventive Problem Solving in Russian). The ARIZ will result in a final solution that the height and angle control parts for a cutting tool should be replaced by a kinematical approach, rather than complicated mechatronic approach(using motors). The Optimize step is composed of two sub-steps: (i) Generating process for obtaining several ideas of piece-removing system by using Brainstorming technique, satisfying the final solution derived from the Design step using ARIZ, and (ii) Optimizing process for selecting the most optimal idea of piece-removing system by using Pugh's matrix from the viewpoints of weight, cost and accuracy. The laststep of Verify has shown that the final design obtained by the 6-Sigma DMADOV technique with ARIZ & Brainstormingcan improve an initial design with design requirements satisfied. In this paper, we have shown that ARIZ and Brainstorming can be cooperatively merged into 6-Sigma DMADOV to give us both a formulatedproblem-solving approach and diverse candidate solutions(or ideas) without trial-and-error efforts.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.31 no.8
• /
• pp.663-671
• /
• 2007

In this work, shape optimization of a wire-wrapped fuel assembly in a liquid metal reactor has been carried out by combining a three-dimensional Reynolds-averaged Navier-Stokes analysis with the radial basis neural network method, a well known surrogate modeling technique for optimization. Sequential Quadratic Programming is used to search the optimal point from the constructed surrogate. Two geometric design variables are selected for the optimization and design space is sampled using Latin Hypercube Sampling. The optimization problem has been defined as a maximization of the objective function, which is as a linear combination of heat transfer and friction loss related terms with a weighing factor. The objective function value is more sensitive to the ratio of the wire spacer diameter to the fuel rod diameter than to the ratio of the wire wrap pitch to the fuel rod diameter. The optimal values of the design variables are obtained by varying the weighting factor.

• Journal of Korean Association for Spatial Structures
• /
• v.23 no.2
• /
• pp.29-36
• /
• 2023

Reinforcement learning (RL) is widely applied to various engineering fields. Especially, RL has shown successful performance for control problems, such as vehicles, robotics, and active structural control system. However, little research on application of RL to optimal structural design has conducted to date. In this study, the possibility of application of RL to structural design of reinforced concrete (RC) beam was investigated. The example of RC beam structural design problem introduced in previous study was used for comparative study. Deep q-network (DQN) is a famous RL algorithm presenting good performance in the discrete action space and thus it was used in this study. The action of DQN agent is required to represent design variables of RC beam. However, the number of design variables of RC beam is too many to represent by the action of conventional DQN. To solve this problem, multi-agent DQN was used in this study. For more effective reinforcement learning process, DDQN (Double Q-Learning) that is an advanced version of a conventional DQN was employed. The multi-agent of DDQN was trained for optimal structural design of RC beam to satisfy American Concrete Institute (318) without any hand-labeled dataset. Five agents of DDQN provides actions for beam with, beam depth, main rebar size, number of main rebar, and shear stirrup size, respectively. Five agents of DDQN were trained for 10,000 episodes and the performance of the multi-agent of DDQN was evaluated with 100 test design cases. This study shows that the multi-agent DDQN algorithm can provide successfully structural design results of RC beam.

• Steel and Composite Structures
• /
• v.44 no.5
• /
• pp.603-617
• /
• 2022

Structural design, in general, is developed through trial and error technique which is guided by standards criteria and based on the intuition and experience of the engineer, a context that leads to structural over-dimensioning, with uneconomic solutions. Aiming to find the optimal design, structural optimization methods have been developed to find a balance between cost, structural safety, and material performance. These methods have become a great opportunity in the steel structural engineering domain since they have as their main purpose is weight minimization, a factor directly correlated to the real cost of the structure. Assuming an objective function of minimum weight with stress and displacement constraints provided by Brazilian standards, the present research proposes the sizing optimization and combined approach of sizing and shape optimization, through a software developed to implement the Simulated Annealing metaheuristic algorithm. Therefore, two steel plane frame layouts, each admitting four typical truss geometries, were proposed in order to expose the difference between the optimal solutions. The assessment of the optimal solutions indicates a notable weight reduction, especially in sizing and shape optimization combination, in which the quantity of design variables is increased along with the search space, improving the efficiency of the optimal solutions achieved.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.14 no.2
• /
• pp.121-126
• /
• 2015

In this research, a tool was developed for the rapid performance of three-dimensional finite element analysis (3D FEA) of a machine tool spindle system made of a shaft and bearings. It runs the FEA with data, such as the bearing stiffness and the coordinates of the points, to define the section of the shaft, bearing positions, and cutting point. developed for the spindle system and then implemented with the tool using an object-oriented programing technique that allows the use of the objects of the CAD system used in this research. Graphic user interfaces were designed for a user to interact with the tool. It provides rapid evaluation of the design of a spindle system, and therefore, it would be helpful to identify a near optimal design of a spindle system based on, say, static stiffness with design changes and, consequently, FEA.

• Korean Journal of Computational Design and Engineering
• /
• v.16 no.6
• /
• pp.397-406
• /
• 2011

Since tolerance allocation in a mobile phone camera manufacturing process greatly affects production cost and reliability of optical performance, a systematic design methodology for allocating optimal tolerances is required. In this study, we proposed the tolerance optimization procedure for determining tolerances that minimize production cost while satisfying the reliability constraints on important optical performance indices. We employed Latin hypercube sampling for evaluating the reliabilities of optical performance and a function-based sequential approximate optimization technique that can reduce computational burden and well handle numerical noise in the tolerance optimization process. Using the suggested tolerance optimization approach, the optimal production cost was decreased by 30.3 % compared to the initial cost while satisfying the two constraints on the reliabilities of optical performance.

• Journal of KSNVE
• /
• v.10 no.3
• /
• pp.437-443
• /
• 2000

The optimal design problem for increasing dynamic stiffness using hybrid model which composed of original detailed BIW(body in white) and impinged beam elements is investigated. Using the characteristics of the beam elements and design sensitivity analysis this approach utilizes an optimization technique to determine the optimal section properties of beam elements. The constraint is to increase the first natural frequency by five percent compared with original one. The results show that the first torsion and bending natural frequencies are increased by five percent using hybrid model and optimization. These results indicate that this optimization method can be employed to enhance the dynamic stiffness of vehicle body structure in design concept stage.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.11b
• /
• pp.599-604
• /
• 2002

Substructures position is considered as design parameter to obtain optimal structural changes to raise its dynamic characteristics. In conventional SDM (structural dynamics modification) method, the layout of modifying substructures position is first fixed and at that condition the structural optimization is performed by using the substructures size and/or material property as design parameters. But in this paper as a design variable substructures global translational and rotational position is treated. For effective structural modification the eigenvalue sensitivity with respect to that design parameter is derived based on measured frequency response function. The optimal structural modification is calculated by combining eigenvalue sensitivities and eigenvalue reanalysis technique iteratively. Numerical examples are presented to the case of beam stiffener optimization to raise the natural frequency of plate.

• IE interfaces
• /
• v.12 no.2
• /
• pp.237-246
• /
• 1999

The Automated Storage/Retrieval System(AS/RS) is very expensive to be built. But there are few tools available which can be used to design the storage systems optimally considering overall system building costs and constraints. In this paper, we propose a design support system for optimal design of AS/RS which includes different sized cells using expert systems and mathematical models considering customer's demands. The optimization models are to be used to minimize the total cost including building, equipment and operation costs. Expert system technique is utilized for using the experiences of skilled designers. By using the proposed system, AS/RS can be designed accurately, economically and efficiently.

• Proceedings of the KIEE Conference
• /
• 1996.11a
• /
• pp.6-8
• /
• 1996

The multiobjective optimization is presented for the optimal design of induction motors. The aim of design is to find an optimized induction motor in terms of both the efficiency and the mass. The efficiency and the mass are linearly combined using the weighting factors. Optimization process is performed by using the improved (1+1) evolution strategy (ES). ES is the algorithm that can find the global minimum. To verify the validity of the proposed method. the method is applied to a sample design.