• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.12 no.2
• /
• pp.173-180
• /
• 2012

This paper proposes a novel optimization algorithm inspired by water flowing and shaking behaviors in a vessel. Water drops in our algorithm flow to the gradient descent direction and are sometimes shaken for getting out of local optimum areas when most water drops fall in local optimum areas. These flowing and shaking operations allow our algorithm to quickly approach to the global optimum without staying in local optimum areas. We experimented our algorithm with four function optimization problems and compared its results with those of particle swarm optimization. Experimental results showed that our algorithm is superior to the particle swarm optimization algorithm in terms of the speed and success ratio of finding the global optimum.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.15 no.7 s.100
• /
• pp.806-812
• /
• 2005

This paper adresses the shaking table tests with 1/100 scaled model followed similitude law for OOdam main designing section to understand nonlinear behavior characteristics of concrete dam body by ground motion. As earthquake wave, Hachinohe and El Centre waves were used and acceleration and displacements are measured to analyze behaviors of dam body. For maximum ground acceleration range $(0.3\~0.9 g)$, the results showed linear behavior regardless of maximum 9round acceleration and secured safety of structure. To analyze the behavior of dam after tension cracking, 3 cm-notch was placed at the critical section of over-flowing section. As results of applying Hachinohe wave(0.8 g), Even though tension cracks were formed at over-flowing section by Hachinohe wave(0.8 g), it showed that the dam is stable for supporting upper stream Part of water tank of dam.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 2005.03a
• /
• pp.110-117
• /
• 2005

This paper deals the shaking table tests with 1/100 scaled model followed by Scott & Iai(1989)'s similitude law for OO dam main desging section to understand nonlinear behavior characteristics of concrete dam body by earthquake wave. As earthquake wave, Hachinohe and Elcentro waves were used and acceleration and displacements are measured to analyze behaviors of dam body. For ground maximum acceleration range ($0.3^{\sim}0.9g$), the results showed linear behavior regardless of ground maximum acceleration and secured safety of structure. To analyze the behavior of dam after tension cracking, 3cm-notch was placed at the critical section of over-flowing section. As results of applying Hachinohe wave(0.8g), Even though tension cracks were formed at over-flowing section by Hachinohe wave(0.8g), it showed that the dam is stable for supporting upper stream part of water tank of dam.

• Journal of Digital Contents Society
• /
• v.18 no.5
• /
• pp.969-976
• /
• 2017

This paper proposes the learning method of an artificial neural network and a convolutional neural network using the WFSO algorithm developed as an optimization algorithm. Since the optimization algorithm searches based on a number of candidate solutions, it has a drawback in that it is generally slow, but it rarely falls into the local optimal solution and it is easy to parallelize. In addition, the artificial neural networks with non-differentiable activation functions can be trained and the structure and weights can be optimized at the same time. In this paper, we describe how to apply WFSO algorithm to artificial neural network learning and compare its performances with error back-propagation algorithm in multilayer artificial neural networks and convolutional neural networks.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.7 no.2
• /
• pp.209-222
• /
• 1997

Authors surveyed the ground water near the waste disposed from a fiberglass production factory to confirm the presence of glassfiber in the water and to determine the effect of sampling conditions and storage on the recovery of fibrous materials in the ground water. Sample was collected at every 4 hours for 48 hours consecutively. After finishing the 48 hours sample, water sampling was done from each tap after repeated turning on and off the water for 30 seconds at each time. Sample was collected in the two 1.5 liter polyethylene bottle after vigorously shaking the bottle with the same water several times with the flowing tap water. At each paired sample, one bottle was stored stand still at room temperature, and the other sample was filtered immediately after sampling. Water was filtered on the Mixed Cellulose Ester filter with negative pressure. Each sample was divided into upper and lower layer. The other bottle was stored at room temperature standstill for 7 days and filtered in the same fashion as the other pair of sample did. Each MCE filter was divided into 4 pieces and one piece was treated with acetone to make it transparent. Each prepared sample was observed by two researchers under the light and polarizing microscopy, scanning electron microscopy and energy dispersive X-ra microanalysis. Fibers were classified by the morphology and polarizing pattern under the polarizing microscope, and count was done. 1. There was a significant fluctuation in number of the fibers, but there was no specific demonstrable pattern. 2. Non-polarizing fibers frequently disappeared after 7 days's storage. But cluster of fibers were found at the wall of the same container by scratching technique. 3. Polarizing fibers were usually found in between the filter and the manicure pasted area. Possible explanations for this phenomenon will be that either these fibers are very light or have electronic polarity. Hence, these fibers are not able to be attached on the surface of slide glass. 4. Under the scanning electron microscopic examination, the fibers which are not refractive under the light microscopy were identified as glassfiber. Other fibers which is refractive under the polarizing microscopy were identified as magnesium silicate fibers. It is strongly suggested that development of standardized method of sample collection and measurement of fibrous material in the water is needed.

• Analytical Science and Technology
• /
• v.12 no.3
• /
• pp.196-202
• /
• 1999

Brown algae, Sargassum horneri, was used as an adsorbent to remove heavy metal Cd(II) and Pb(II) ions. Sargassum horneri picked in East Coast was formulated into the grain with the size of 40-60 mesh after wind dry. Batch method was used in order to investigate the adsorption rate by measuring the adsorption amounts with shaking time. In the column method, the adsorption amounts were measured by flowing metal solutions into the algae-packed column at the rate of 1 ml/min. Adsorption amounts in both batch method and column method were in the following order : pH 10.5>7.0>3.5. It was found that Pb(II) was more adsorbed on the algae grain than Cd(II). It was also revealed that the adsorption amounts reached the maximum within 5 minutes irrespective of pH condition in the batch method. It was concluded that the batch method was more effective than the column method in terms of recovery rate.