• Structural Engineering and Mechanics
• /
• v.82 no.3
• /
• pp.271-282
• /
• 2022

The active learning surrogate model based on adaptive sampling strategy is increasingly popular in reliability analysis. However, most of the existing sampling strategies adopt the trial and error method to determine the size of the Monte Carlo (MC) candidate sample pool which satisfies the requirement of variation coefficient of failure probability. It will lead to a reduction in the calculation efficiency of reliability analysis. To avoid this defect, a new method for determining the optimal size of the MC candidate sample pool is proposed, and a new structural reliability analysis method combining polynomial chaos-based Kriging model (PC-Kriging) with adaptive sampling region is also proposed (PCK-ASR). Firstly, based on the lower limit of the confidence interval, a new method for estimating the optimal size of the MC candidate sample pool is proposed. Secondly, based on the upper limit of the confidence interval, an adaptive sampling region strategy similar to the radial centralized sampling method is developed. Then, the k-means++ clustering technique and the learning function LIF are used to complete the adaptive design of experiments (DoE). Finally, the effectiveness and accuracy of the PCK-ASR method are verified by three numerical examples and one practical engineering example.

• Transactions of Materials Processing
• /
• v.29 no.2
• /
• pp.89-96
• /
• 2020

The finite element method has been widely applied in the sheet metal forming process. However, the finite element method is computationally expensive and time consuming. In order to tackle this problem, surrogate modeling methods have been proposed. An artificial neural network (ANN) is one such surrogate model and has been well studied over the past decades. However, when it comes to ANN with two or more layers, so called deep neural networks (DNN), there is distinct a lack of research. We chose to use DNNs our surrogate model to predict the behavior of sheet metal in the deep drawing process. Thickness variation is selected as an output of the DNN in order to evaluate workpiece feasibility. Input variables of the DNN are radius of die, die corner and blank holder force. Finite element analysis was conducted to obtain data for surrogate model construction and testing. Sampling points were determined by full factorial, latin hyper cube and monte carlo methods. We investigated the performance of the DNN according to its structure, number of nodes and number of layers, then it was compared with a radial basis function surrogate model using various sampling methods and numbers. The results show that our DNN could be used as an efficient surrogate model for the deep drawing process.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.21 no.5
• /
• pp.8-14
• /
• 2013

In this study, we propose the method of optimally selecting material of front pillar (A-pillar) and package tray for minimizing weight while satisfying vehicle requirements on static stiffness and dynamic stiffness. First, we formulate a material selection optimization problem. Next, we establish the CAE procedure of evaluating static stiffness and dynamic stiffness. Then, to enhance the efficiency of design work, we integrate and automate the established CAE procedure using a commercial process integration and design optimization (PIDO) tool, PIAnO. For effective optimization, we adopt the approach of metamodel based approximate optimization. As a sampling method, an orthogonal array (OA) is used for selecting sampling points. The response values are evaluated at the sampling points and then these response values are used to generate a metamodel of each response using the radial basis function regression (RBFr). Using the RBFr models, optimization is carried out an evolutionary algorithm that can handle discrete design variables. Material optimization result reveals that the weight is reduced by 49.8% while satisfying all the design constraints.

• Korean Journal of Bronchoesophagology
• /
• v.17 no.1
• /
• pp.9-13
• /
• 2011

Endobronchial ultrasound (EBUS), which enables visualization of lesions beyond the bronchus, broadens the fields of bronchoscopy. Two types of ultrasound, radial and linear, are used for bronchoscopy. Radial EBUS is performed by inserting an ultrasound mini-probe through the working channel of a flexible bronchoscope. Evaluation of the depth of invasion of early endobronchial lung cancers using radial EBUS is useful in deciding endobronchial treatment. A central tumor limited to within the cartilaginous layer is a good indication for endobronchial photodynamic therapy. EBUS-guide sheath (GS) technique is a sampling method assisted by localization of peripheral lesions using EBUS. The diagnostic yield of EBUS-GS method is higher than that of conventional transbronchial biopsy. High diagnostic values of EBSU-GS method are reported even in small (${\leq}2cm$) peripheral tumors. Linear EBUS is used for endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA). EBUS-TBNA has high diagnostic yields in mediastinal staging of lung cancer even in patients having radiologically early stage lung cancers with normal CT or PET findings in the mediastinum. EBUS is a valuable method in evaluating early endobronchial tumors and peripheral small lung cancers and as well as in mediastinal staging.

• The KSFM Journal of Fluid Machinery
• /
• v.12 no.4
• /
• pp.44-53
• /
• 2009

Numerical design optimization of a fan-shaped hole for film-cooling has been carried out to improve film-cooling effectiveness 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. The injection angle of hole, lateral expansion angle of hole and ratio of length-to-diameter of the hole are chosen as design variables and spatially averaged film-cooling effectiveness is considered as an objective function which is to be maximized. Twenty training points are obtained by Latin Hypercube sampling for three design variables. Sequential quadratic programming is used to search for the optimal point from the constructed surrogate. The film-cooling effectiveness has been successfully improved by the optimization with increased value of all design variables as compared to the reference geometry.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.4
• /
• pp.735-741
• /
• 2009

This paper presents an algorithm for the permanent magnet shape optimization of a large scale BLDC(Brushless DC) motor to minimize the cogging torque. A response surface method (RSM) using multiquadric radial basis function is employed to interpolate the objective function in design parameter space. In order to get a reasonable response surface with relatively small number of sampling data points, additional sampling points are added on the basis of design sensitivity analysis computed by using FEM. The algorithm has 2 stages: the first stage is to determine the PM arc angle, and the 2nd stage is to optimize the magnet pole shape. The developed algorithm is applied to a 5MW BLDC motor to get a minimum cogging torque. After 3 iterations with 4 design parameters, the cogging torque is reduced to 13.2% of the initial one.

• International Journal of Precision Engineering and Manufacturing
• /
• v.9 no.4
• /
• pp.45-50
• /
• 2008

The Fourier-Mellin transform is the theoretical basis for the translation, rotation, and scale invariance of an image. However, its implementation requires a log-polar map of the original image, which requires logarithmic sampling of a radial variable in that image. This means that the mapping process is accompanied by considerable loss of data. To solve this problem, we propose a dual log-polar map that uses both a forward image map and a reverse image map simultaneously. Data loss due to the forward map sub-sampling can be offset by the reverse map. This is the first step in creating an invertible log-polar map. Experimental results have demonstrated the effectiveness of the proposed scheme.

• Journal of Electrical Engineering and Technology
• /
• v.2 no.2
• /
• pp.152-156
• /
• 2007

This paper proposes a method for real-time control of both capacitors and ULTC in a distribution system to reduce the total power loss and to improve the voltage profile over the course of a day. The multi-stage consists of the off-line stage to determine dispatch schedule based on a load forecast and the on-line stage generates the time and control sequences at each sampling time. It is then determined whether one of the control actions in the control sequence is performed at the present sampling time. The proposed method is presented for a typical radial distribution system with a single ULTC and capacitors.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.32 no.4
• /
• pp.359-370
• /
• 2022

Objective: This study is intended to design a commercially available passive sampler and conduct performance test on its use as a media for evaluating a working environment. Methods: This study was conducted to select adsorbents, design models, and evaluate storage stability and sampling rates for the development of new types of passive samplers. Results: The impurity detection, adsorbent capacity and breakthrough volume of five types of activated carbon were tested for selection of an adsorbent. One product was selected in consideration of the efficiency of purchase. A number of passive samplers were designed in a radial style and a badge style using plastic as a material. The final two prototypes were made using molds or 3D printing. For the storage stability evaluation, samples were stored at different temperature for 1~21 days and then analyzed. Most of the chemicals had excellent storage stability when refrigerated. However, some chemicals such as dichloromethane and methyl ethyl ketone need to be analyzed as soon as possible after sampling. Conclusion: In this study, new types of passive samplers for 66 chemical compounds were developed. The evaluation of storage stability and sampling rates showed different results depending on the properties of the chemical substance. For some chemicals such as methyl ethyl ketone and dimethylformamide, activated carbon is inappropriate as an absorbent. In future studies, additional experiments are required on chemicals that are difficult to collect with activated carbon.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.18 no.3
• /
• pp.662-669
• /
• 1994

In the combustion system, the optimum spray conditions reduce the pollutant emission of exhaust gas and enhance the fuel efficiency. The spray characteristics-the drop size, the drop velocity, the number density and the mass flux, become increasingly important in the design of combustor and in testifying numerical simulation of spray flow in the combustor. The purposes of this study are to clarify the spray characteristics of twin-fluid nozzle and to offer the data for combustor design and the numerical simulation of a spray flow. Spatial drop diameter was measured by immersion sampling method. The mean diameter, size distribution and uniformity of drop were analyzed with variations of air/liquid mass flow ratio. The results show that the SMD increases with the liquid supply flow rate and decreases with the air supply velocity. The radial distribution of SMD shows the larger drops can diffuse farther to the boundary of spray. And the drop size range is found to be wider close to the spray boundary where the maximum SMD locates.