• Proceedings of the KSME Conference
• /
• 2001.06b
• /
• pp.584-589
• /
• 2001

This paper presents the dynamic parameter design optimization of a suspension seat system using the genetic algorithm. At first, an equivalent 1-D.O.F. mass-spring-damper model of a suspension seat system was constructed for the purpose of its vibration analysis. Vertical vibration response and transmissibility of the equivalent model due to base excitations, which are defined in the ISO's seat vibration test codes, were computed. Furthermore, seat vibration test, that is ISO's damping test, was carried out in order to investigate the validity of the equivalent suspension seat model. Both analytical and experimental results showed good agreement each other. For the design optimization, the acceleration transmissibility of the suspension seat model was adopted as an object function. A simple genetic algorithm was used to search the optimum values of the design variables, suspension stiffness and damping coefficient. Finally, vibration ride performance test results showed that the optimum suspension parameters gives the lowest vibration transmissibility. Accordingly the genetic algorithm and the equivalent suspension seat modelling can be successfully adopted in the vibration ride quality optimization of a suspension seat system.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.28 no.7
• /
• pp.879-886
• /
• 2004

A numerical optimization has been carried out to determine the shape of the three-dimensional channel with oblique ribs attached on both walls to enhance turbulent heat transfer. The response surface based optimization is used as an optimization technique with Reynolds-averaged Navier-Stokes analysis of fluid flow and heat transfer. Shear stress transport (SST) turbulence model is used as a turbulence closure. Numerical results fur heat transfer rate show good agreements with experimental data. four dimensionless variables such as, rib pitch-to-rib height ratio, rib height-to-channel height ratio, streamwise rib distance on opposite wall to rib pitch ratio, and the attack angle of the rib are chosen as design variables. The objective function is defined as a linear combination of heat-transfer and friction-loss related coefficients with a weighting factor. D-optimal method is used to determine the training points as a means of design of experiment. Sensitivity of the objective parameters to each design variable has been analyzed. And, optimal values of the design variables have been obtained in a range of the weighting factor.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.18 no.12
• /
• pp.1217-1223
• /
• 2008

Vehicle vibration mostly originates from the road excitation and causes discomfort, fatigue and even injury to a driver. Vehicle ride comfort is one of the most important performance indices to achieve a high-quality vehicle design. Since design parameter variations inevitably result in the vehicle ride comfort variance, the variance characteristics should be analyzed in the early design stage of the vehicle. The vehicle ride comfort is often defined by an index which employs a weighted RMS value of the acceleration PSD of a seat position. The design solution is obtained through two steps in this study. An optimization problem to obtain a minimum ride comfort index is solved first. Then another optimization problem to obtain minimum variance of the ride comfort index is solved. For the optimization problems, the equations of motion and the sensitivity equations are derived basing on a 5-DOF vehicle model. The numerical results show that an optimal solution for the minimum ride comfort is not necessarily same as that of the minimum variance of the ride comfort.

• Smart Structures and Systems
• /
• v.18 no.3
• /
• pp.425-448
• /
• 2016

In this paper, a new Pigeon Colony Algorithm (PCA) based on the features of a pigeon colony flying is proposed for solving global numerical optimization problems. The algorithm mainly consists of the take-off process, flying process and homing process, in which the take-off process is employed to homogenize the initial values and look for the direction of the optimal solution; the flying process is designed to search for the local and global optimum and improve the global worst solution; and the homing process aims to avoid having the algorithm fall into a local optimum. The impact of parameters on the PCA solution quality is investigated in detail. There are low-dimensional functions, high-dimensional functions and systems of nonlinear equations that are used to test the global optimization ability of the PCA. Finally, comparative experiments between the PCA, standard genetic algorithm and particle swarm optimization were performed. The results showed that PCA has the best global convergence, smallest cycle indexes, and strongest stability when solving high-dimensional, multi-peak and complicated problems.

• Earthquakes and Structures
• /
• v.7 no.3
• /
• pp.271-294
• /
• 2014

In recent years, along with the advances made in performance-based design optimization, the need for fast calculation of response parameters in dynamic analysis procedures has become an important issue. The main problem in this field is the extremely high computational demand of time-history analyses which may convert the solution algorithm to illogical ones. Two simplifying strategies have shown to be very effective in tackling this problem; first, simplified nonlinear modeling investigating minimum level of structural modeling sophistication, second, wavelet analysis of earthquake records decreasing the number of acceleration points involved in time-history loading. In this paper, we try to develop an efficient framework, using both strategies, to solve the performance-based multi-objective optimal design problem considering the initial cost and the seismic damage cost of steel moment-frame structures. The non-dominated sorting genetic algorithm (NSGA-II) is employed as the optimization algorithm to search the Pareto optimal solutions. The constraints of the optimization problem are considered in accordance with Federal Emergency Management Agency (FEMA) recommended design specifications. The results from numerical application of the proposed framework demonstrate the capabilities of the framework in solving the present multi-objective optimization problem.

• Steel and Composite Structures
• /
• v.46 no.1
• /
• pp.33-51
• /
• 2023

This research presents a multi-material topology optimization for functionally graded material (FGM) and nonFGM with elastic buckling criteria. The elastic buckling based multi-material topology optimization of functionally graded steels (FGSs) uses a Jacobi scheme and a Method of Moving Asymptotes (MMA) as an expansion to revise the design variables shown first. Moreover, mathematical expressions for modified interpolation materials in the buckling framework are also described in detail. A Solid Isotropic Material with Penalization (SIMP) as well as a modified penalizing material model is utilized. Based on this investigation on the buckling constraint with homogenization material properties, this method for determining optimal shape is presented under buckling constraint parameters with non-homogenization material properties. For optimal problems, minimizing structural compliance like as an objective function is related to a given material volume and a buckling load factor. In this study, conflicts between structural stiffness and stability which cause an unfavorable effect on the performance of existing optimization procedures are reduced. A few structural design features illustrate the effectiveness and adjustability of an approach and provide some ideas for further expansions.

• International conference on construction engineering and project management
• /
• 2020.12a
• /
• pp.225-234
• /
• 2020

Reliability-based structural optimization usually requires designers or engineers model different designs manually, which is considered very time consuming and all possibilities cannot be fully explored. Otherwise, a lot of time are needed for designers or engineers to learn mathematical modeling and programming skills. Therefore, a framework that integrates generative design, structural simulation and reliability theory is proposed. With the proposed framework, various designs are generated based on a set of rules and parameters defined based on visual programming, and their structural performance are simulated by OpenSees. Then, reliability of each design is evaluated based on the simulation results, and an optimal design can be found. The proposed framework and prototype are tested in the optimization of a steel frame structure, and results illustrate that generative design based on visual programming is user friendly and different design possibilities can be explored in an efficient way. It is also reported that structural reliability can be assessed in an automatic way by integrating Dynamo and OpenSees. This research contributes to the body of knowledge by providing a novel framework for automatic reliability evaluation and structural optimization.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.6
• /
• pp.319-325
• /
• 2023

In this study, we proposed a method for hyperparameter optimization in the building and training of a deep learning model designed to process point cloud data collected by a millimeter-wave radar system. The primary aim of this study is to facilitate the deployment of a baseline model in resource-constrained IoT devices. We evaluated a RadHAR baseline deep learning model trained on a public dataset composed of point clouds representing five distinct human activities. Additionally, we introduced a coarse-to-fine hyperparameter optimization procedure, showing substantial potential to enhance model efficiency without compromising predictive performance. Experimental results show the feasibility of significantly reducing model size without adversely impacting performance. Specifically, the optimized model demonstrated a 3.3% improvement in classification accuracy despite a 16.8% reduction in number of parameters compared th the baseline model. In conclusion, this research offers valuable insights for the development of deep learning models for resource-constrained IoT devices, underscoring the potential of hyperparameter optimization and model size reduction strategies. This work contributes to enhancing the practicality and usability of deep learning models in real-world environments, where high levels of accuracy and efficiency in data processing and classification tasks are required.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.15 no.1
• /
• pp.27-33
• /
• 2005

In this study, the process parameters in ceramics powder compaction are optimized for getting high relative densities of ceramic products. To find optimized parameters, the analytic models of powder compaction are firstly prepared by 2-dimensional rod arrays with random green densities using a quasi-random multiparticle array. Then, using finite element method, the changes in relative densities are analyzed by varying the size of Al₂O₃ particle, the amplitude of cyclic compaction, and the coefficient of friction, which influence the relative density in cyclic compactions. After the analytic function of relative density associated process parameters are formulated by aid of the response surface method, the optimal conditions in powder compaction process are found by the grid search method. When the particle size of Al₂O₃ is 22.5 ㎛, the optimal parameters for the amplitude of cyclic compaction and the coefficient of friction are 75 MPa and 0.1103, respectively. The maximum relative density is 0.9390.

• Journal of KIISE:Software and Applications
• /
• v.37 no.9
• /
• pp.694-705
• /
• 2010

The mutation operation is the main operation in the evolutionary programming which has been widely used for the optimization of real valued function. In general, the mutation operation utilizes both a probability distribution and its parameter to change values of variables, and the parameter itself is subject to its own mutation operation which requires other parameters. However, since the optimal values of the parameters entirely depend on a given problem, it is rather hard to find an optimal combination of values of parameters when there are many parameters in a problem. To solve this shortcoming at least partly, if not entirely, in this paper, we propose a new mutation operation in which the parameter for the variable mutation is theoretically estimated from the self-adaptive perspective. Since the proposed algorithm estimates the scale parameter of the Cauchy probability distribution for the mutation operation, it has an advantage in that it does not require another mutation operation for the scale parameter. The proposed algorithm was tested against the benchmarking problems. It turned out that, although the relative superiority of the proposed algorithm from the optimal value perspective depended on benchmarking problems, the proposed algorithm outperformed for all benchmarking problems from the perspective of the computational time.