• Journal of the Institute of Electronics Engineers of Korea CI
• /
• 제39권6호
• /
• pp.47-52
• /
• 2002

The wavelet functions are originated from scaling functions and can be used as activation function in the hidden node of the network by deciding two parameters such as scale and center. In this paper, we would like to propose the mixed structure. When we compose the WNN using wavelet functions, we propose to set a single scale function as a node function together. The properties of the proposed structure is that while one scale function approximates the target function roughly, the other wavelet functions approximate it finely. During the determination of the parameters, the wavelet functions can be determined by the global search algorithm such as genetic algorithm to be suitable for the suggested problem. Finally, we use the back-propagation algorithm in the learning of the weights.

• Journal of the Society of Naval Architects of Korea
• /
• 제59권1호
• /
• pp.39-45
• /
• 2022

Normal strategy of structure optimization procedure has been minimum cost or weight design. Minimum weight design satisfying an allowable stress has been used for the ship and offshore structure, but minimum cost design could be used for the case of high human cost. Natural frequency analysis and forced vibration one have been used for the strength estimation of marine structures. For the case of high precision experiment facilities in marine field, the structure has normally enough margin in allowable stress aspect and sometimes needs high natural frequency of structure to obtain very high precise experiment results. It is not easy to obtain a structure design with high natural frequency, since the natural frequency depend on the stiffness to mass ratio of the structure and increase of structural stiffness ordinary accompanies the increase of mass. It is further difficult at the grillage structure design using the profiles, because the properties of profiles are not continuous but discrete, and resource of profiles are limited at the design of grillage structure. In this paper, the grillage structure design system under the constraint of high natural frequency is introduced. The design system adopted genetic algorithm to realize optimization procedure and can be used at the design of the experimental facilities of marine field such as a towing carriage, PMM, test frame, measuring frame and rotating arm.

• Journal of the Korean Society of Industry Convergence
• /
• 제26권1호
• /
• pp.113-119
• /
• 2023

In this paper, we propose an optimization of a pose estimation deep learning model for recognition of abnormal behavior in unmanned stores using radio frequencies. The radio frequency use millimeter wave in the 30 GHz to 300 GHz band. Due to the short wavelength and strong straightness, it is a frequency with less grayness and less interference due to radio absorption on the object. A millimeter wave radar is used to solve the problem of personal information infringement that may occur in conventional CCTV image-based pose estimation. Deep learning-based pose estimation models generally use convolution neural networks. The convolution neural network is a combination of convolution layers and pooling layers of different types, and there are many cases of convolution filter size, number, and convolution operations, and more cases of combining components. Therefore, it is difficult to find the structure and components of the optimal posture estimation model for input data. Compared with conventional millimeter wave-based posture estimation studies, it is possible to explore the structure and components of the optimal posture estimation model for input data using genetic algorithms, and the performance of optimizing the proposed posture estimation model is excellent. Data are collected for actual unmanned stores, and point cloud data and three-dimensional keypoint information of Kinect Azure are collected using millimeter wave radar for collapse and property damage occurring in unmanned stores. As a result of the experiment, it was confirmed that the error was moored compared to the conventional posture estimation model.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• 제41권3호
• /
• pp.173-184
• /
• 2013

A multi-level design optimization framework for aerodynamic design of rotary wing such as propeller and helicopter rotor blades is presented in this study. Strategy of the proposed framework is to enhance aerodynamic performance by sequentially applying the planform and sectional design optimization. In the first level of a planform design, we used a genetic algorithm and blade element momentum theory (BEMT) based on two-dimensional aerodynamic database to find optimal planform variables. After an initial planform design, local flow conditions of blade sections are analyzed using high-fidelity CFD methods. During the next level, a sectional design optimization is conducted using two dimensional Navier-Stokes analysis and a gradient based optimization algorithm. When optimal airfoil shape is determined at the several spanwise locations, a planform design is performed again. Through this iterative design process, not only an optimal flow condition but also an optimal shape of an EAV propeller blade is obtained. To validate the optimized propeller-blade design, it is tested in wind-tunnel facility with different flow conditions. An efficiency, which is slightly less than the expected improvement of 7% predicted by our proposed design framework but is still satisfactory to enhance the aerodynamic performance of EAV system.