• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2007년도 춘계학술대회A
• /
• pp.933-938
• /
• 2007

These days, up-to-date humanoid robots are continuously developed. Among them, Qrio, Asimo[1,2] are famous for its unique walking technology and natural movement. These robots could show manufacturers' technological improvement and leave a good impression to the customer. In accordance with global trends, Samsung is also producing humanoid robot. The humanoid robot, however, could walk like a human compared to the industrial robot fixed in the factory. This feature could cause another dynamic effect while walking. In this paper, the robot's feet were examined to find out parameters that affect stability of the humanoid robot's feet. With the sensitivity analysis, the optimization procedure in design of experiments finds the most suitable performance of robot. Maximum deflection of the frame upon various cases was minimized, and rubber coefficients for shock absorption were optimized.

• 한국멀티미디어학회논문지
• /
• 제21권4호
• /
• pp.441-449
• /
• 2018

Deep learning has been used for various processing centered on image recognition. One core algorithms of the deep learning, convolutional neural network is an deep neural network that specialized in image recognition. In this paper, we use a convolutional neural network to classify forest insects and propose an optimization method. Experiments were carried out by combining two weight initialization and six parameter update methods. As a result, the Xavier-SGD method showed the highest performance with an accuracy of 82.53% in the 12 different combinations of experiments. Through this, the latest learning algorithms, which complement the disadvantages of the previous parameter update method, we conclude that it can not lead to higher performance than existing methods in all application environments.

• 음성과학
• /
• 제15권4호
• /
• pp.85-96
• /
• 2008

Automatic speech recognition has severe performance degradation under noisy environments. To cope with the noise problem, many methods have been proposed. Most of them focused on noise-robust features or model adaptation. However, researchers have overlooked utterance verification (UV) under noisy environments. In this paper we discuss UV problems based on the normalized confidence measure. First, we show that UV performance is also degraded in noisy environments with the experiments of an isolated word recognition. Then we observe how the degradation of UV performances is suffered. Based on the UV experiments we propose a modeling method of the statistics of phone confidences using sigmoid functions. For obtaining the parameters of the sigmoidal models, the particle swarm optimization (PSO) is adopted. The proposed method improves 20% rejection performance. Our experimental results show that the PSO-NCM can apply noise speech recognition successfully.

• ETRI Journal
• /
• 제43권3호
• /
• pp.549-560
• /
• 2021

Cyberattacks are often difficult to identify with traditional signature-based detection, because attackers continually find ways to bypass the detection methods. Therefore, researchers have introduced artificial intelligence (AI) technology for cybersecurity analysis to detect malicious PowerShell scripts. In this paper, we propose a feature optimization technique for AI-based approaches to enhance the accuracy of malicious PowerShell script detection. We statically analyze the PowerShell script and preprocess it with a method based on the tokens and abstract syntax tree (AST) for feature selection. Here, tokens and AST represent the vocabulary and structure of the PowerShell script, respectively. Performance evaluations with optimized features yield detection rates of 98% in both machine learning (ML) and deep learning (DL) experiments. Among them, the ML model with the 3-gram of selected five tokens and the DL model with experiments based on the AST 3-gram deliver the best performance.

• 한국정밀공학회지
• /
• 제20권2호
• /
• pp.168-175
• /
• 2003

In this research, a bridge-type flexure hinge mechanism is developed and optimized to amplify the displacement of a multilayer piezostack. Developed hinge mechanism has three-dimensional structure to reduce link size, so it have high amplification ratio with respect to small size. A flexure hinge is assumed to be 6 degree-of-freedom spring elements and matrix methods are used to model a hinge mechanism. To verify derived matrix model, a displacement and frequency experiments are performed. The analysis result shows that the displacemental error between matrix model and experiments is below 10 percents and the deformation of hinge in parasitic direction should be considered In hinge modeling. Using developed matrix model, an optimal design is performed to maximize the performance of hinge mechanism.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2014년도 추계학술대회 논문집
• /
• pp.743-744
• /
• 2014

Hyperbolic metamaterials in which waves can only propagate through the radial direction have achieved much attention these days due to their capability of sub-wavelength resolution. In this work, the realization and optimization of hyperbolic elastic metamaterials are mainly studied. To obtain a new hyperbolic elastic metamaterial, a specially-engineered mass-spring system is introduced. Based on the mass-spring system, the hyperbolic elastic metamaterials are proposed and realized. In addition, the sub-wavelength resolution of the proposed hyperbolic elastic metamaterial is verified by ultrasonic elastic wave experiments. For the experiments, specially-designed magnetostrictive patch transducers are developed to realize two sub-wavelength elastic wave sources. Furthermore, the proposed hyperbolic elastic metamaterial is optimized to maximize its operating frequency ranges by the topology optimization method.

• International Journal of Fluid Machinery and Systems
• /
• 제2권1호
• /
• pp.1-12
• /
• 2009

Surrogate modeling is applied to a compressor blade shape optimization to modify its stacking line and thickness to enhance adiabatic efficiency and total pressure ratio. Six design variables are defined by parametric curves and three objectives; efficiency, total pressure and a combined objective of efficiency and total pressure are considered to enhance the performance of compressor blade. Latin hypercube sampling of design of experiments is used to generate 55 designs within design space constituted by the lower and upper limits of variables. Optimum designs are found by formulating a PRESS (predicted error sum of squares) based averaging (PBA) surrogate model with the help of a gradient based optimization algorithm. The optimum designs using the current variables show that, to optimize the performance of turbomachinery blade, the adiabatic efficiency objective is improved substantially while total pressure ratio objective is increased a very small amount. The multi-objective optimization shows that the efficiency can be increased with the less compensation of total pressure reduction or both objectives can be increased simultaneously.

• Design & Manufacturing
• /
• 제16권1호
• /
• pp.62-69
• /
• 2022

In the present study, the injection molding optimization of a large thick LH type elastic frames for the reduction of warpage was performed. Two kinds of fine and coarse finite element models were prepared to investigate the efficiency of analysis time and quality on simulation results. In order to derive injection molding conditions that can minimize distortion of parts, it was investigated that the effects of mold temperature, resin temperature, injection time, hold pressure switching time, holding pressure and the hold time on deformation characteristics using the design of experiments. The main influential factors on the warpage were found from the optimization simulation and the geometry data of the warpage result was converted into an initial model for injection simulation. It was shown that a coarse model with good mesh quality could be adapted for mold design since the total analysis time using the proposed model was reduced to 1/10. The suggested inversed warpage model produced the best minimized result of warpage.

• 한국유체기계학회 논문집
• /
• 제17권3호
• /
• pp.5-12
• /
• 2014

This paper presents a design optimization for meridional profile and blade angle ${\theta}$ of a centrifugal compressor with DOE (design of experiments) and RSM (response surface method). Control points of the $3^{rd}$ order Bezier curve are used for design parameters and specific overall efficiency is used as object function. The response surface function shows good agreement with the 3D computational results. Three different optimized designs are proposed and compared with reference design at design point and off-design point. Contours of relative Mach number, static entropy, and total pressure are analyzed for improvement of performance by optimization. Off-design performance analysis is conducted by total pressure and efficiency.

• Journal of applied mathematics & informatics
• /
• 제26권3_4호
• /
• pp.811-818
• /
• 2008

Biological systems with both protein-protein and protein-gene interactions can be modeled by differential equations for concentrations of the proteins with time-delay terms because of the time needed for DNA transcription to mRNA and translation of mRNA to protein. Values of some parameters in the mathematical model can not be measured owing to the difficulty of experiments. Also values of some parameters obtained in a normal stress condition can be changed under pathological stress stimuli. Thus it is important to find the effective way of determining parameters values. One approach is to use optimization algorithms. Here we construct an optimal system used to find optimal parameters in the equations with nonnegative time delays and apply this optimization result to the Nuclear factor-${\kappa}B$ pathway.