• 한국환경과학회지
• /
• 제10권5호
• /
• pp.373-380
• /
• 2001

In recent years, the demand of activated carbon has been increasing steadily because of the environmental problems. Among them waste and water treatment and removal of poisonous gas were invorved. Therefore, in this study, activated carbon was made from the waste chinese cabbage and measured the iodine adsorption ability, carbonization yield, and activation yield of the produced activated carbon. The result showed that the carbonization yield was decreased when carbonization temperature was increaed and that the optimal carbonization temperature was $600{\circ}C$. The optimal concentration of NaOH for removing ash in the raw sample was 1~2N. The range of iodine adsorption number of activated carbon using chinese cabbage at $600{\circ}C$. carbonization was 610.82mg/g to 1019.58mg/g. The activation result of carbonization sample showed that the optimal activation condition was the carbonization at $400\circ}C$ and the activation at$700{\circ}C$. So the production of activated carbon using chinese cabbage was possible in the aspect of reuse of resource and decrease of environmental pollution compared to the commercial activated carbon.

• 한국인공지능학회지
• /
• 제5권1호
• /
• pp.29-37
• /
• 2017

The range of problems that can be handled by the activation of big data and the development of hardware has been rapidly expanded and machine learning such as deep learning has become a very versatile technology. In this paper, mnist data set is used as experimental data, and the Cross Entropy function is used as a loss model for evaluating the efficiency of machine learning, and the value of the loss function in the steepest descent method is We applied the Gradient Descent Optimize algorithm to minimize and updated weight and bias via backpropagation. In this way we analyze optimal reliability value corresponding to the number of exercises and optimal reliability value without overfitting. And comparing the overfitting time according to the number of data changes based on the number of training times, when the training frequency was 1110 times, we obtained the result of 92%, which is the optimal reliability value without overfitting.

• 한국경영과학회지
• /
• 제38권3호
• /
• pp.13-22
• /
• 2013

Queueing system with scheduled arrivals is useful in many fields where both customers' waiting time and servers' operation time (utilization) are important, and arrival time of customers is possible to be controlled. In this paper, we analyzed the operation of ship maintenance with the queueing system with scheduled arrival. Based on the model presented by Pegden and Rosenshine [8], who dealt with exponential service time, we extended the service time distributions to phase-type distribution which is able to include a wide range of real stochastic phenomena. Since most activities in the military are carried out under tight control and schedule, scheduled arrival queue has quite good applicability in this area. In this context, we applied queue with scheduled arrival to the optimal booking time decision for the ship maintenance in the navy.

• 해양환경안전학회지
• /
• 제1권1호
• /
• pp.9-25
• /
• 1995

Although the studies carried out in recent years have provied much new information about channel widths and alignment, they are not consistent in their results. In addition, as a result of variations in local condition and type of traffic accommodated, the dimensions of the channel widths vary over a wide range. Therefore, the recommendation made by the maritime engineering organizations over the world, do not offer detailed and decisive optimal design criteria and are all different. It, therefore, was attempted in this paper to draw a decisive guideline on the optimal widths and alignment of the navigational channels, which can be utilized by the port designers at the stage of the planning. The guideline was drawn through the comparison and analysis of the existing guidelines of the U.S.A, Japan and PIANC and simulation experiment. The simulation experiment was carried out using the "Off Line Port and Waterway Design Simulator" to find the optimal dimensions of the widths of the navigational channels. 90 different simulation runs were conducted at the 3 different secenario channels. New guidelines, the result of the study, is expected to be used usefully by the Korean port designers when designing the rapodly developing ports in Korea. in Korea.

• 한국물환경학회지
• /
• 제25권5호
• /
• pp.758-763
• /
• 2009

This study presents the optimal conditions of zero-valent iron (ZVI) pretreatment for color removal from real dye wastewater. Removal of color by ZVI was strongly subject to the acidity of the wastewater buffering the pH increased after ZVI reduction. The real dye wastewater did not contain a sufficient amount of acidity and thus it was necessary to supplement acid to the dye wastewater before treatment. In continuous operation of iron column, the empty bed contact time (EBCT) and initial pH were varied to find the optimal conditions. A non-linear regression model fitted well the experimental result predicting that the optimal EBCT and pH for 80% removal efficiency was present in the range of 57~90 and 5~5.9, respectively. Color of column effluents could be further removed in the following biological oxidation step and the biodegradability of wastewater was also enhanced after iron pretreatment.

• 소성∙가공
• /
• 제15권4호
• /
• pp.304-310
• /
• 2006

Due to a recent growth of the area of MEMS and a trend moving toward smaller scale, a micro manufacturing that is usually related with lithography is now emerging. Differently from traditional manufacturing processes, the micro or miniature manufacturing usually requires expensive sophisticated equipments and its characteristics are of high cost and of low productivity. However, a miniature stamping, which makes small sized product with a thin metal usually in the range of meso-scale, can be realized in a low cost and in a high productivity with relatively inexpensive equipments. For a successful development of miniature stamping, lots of obstacles, including material properties related with formability, have to be overcome. Since the thin metal shows distinctive characteristics, e.g., size effect and statistically scattered material properties, the formability of miniature stamping is not good in general and the possible shape with the miniature stamping is limited relatively simple shapes. Since the optimal blank improves formability and the improved formability can make up for problems of material properties, the possibility of success can be increased. This study is carried out to show the possibility of miniature stamping and to verify the effect of optimal blank for the miniature stamping.

• International Journal of Advanced Culture Technology
• /
• 제5권2호
• /
• pp.82-89
• /
• 2017

The range of problems that can be handled by the activation of big data and the development of hardware has been rapidly expanded and machine learning such as deep learning has become a very versatile technology. In this paper, mnist data set is used as experimental data, and the Cross Entropy function is used as a loss model for evaluating the efficiency of machine learning, and the value of the loss function in the steepest descent method is We applied the GradientDescentOptimize algorithm to minimize and updated weight and bias via backpropagation. In this way we analyze optimal reliability value corresponding to the number of exercises and optimal reliability value without overfitting. And comparing the overfitting time according to the number of data changes based on the number of training times, when the training frequency was 1110 times, we obtained the result of 92%, which is the optimal reliability value without overfitting.

• 한국정밀공학회지
• /
• 제19권12호
• /
• pp.163-170
• /
• 2002

The quality of a precision product, in general, relies on the accuracy and precision of its manufacturing and inspection process. In many cases, the level of precision in the manufacturing and inspection system is also dependent on the positioning capability of tool with respect to the work piece in the process. Recently the positioning accuracy level has reached to the level of submicron and long range of motion is required. For example, for 1 GDARM lithography, 20nm accuracy and 300mm stroke needs. This paper refers to the lithography stage especially to fine stage. In this study, for long stroke and high accuracy, the dual servo system is proposed. For the coarse actuator, LDM (Linear DC Motor) is used and for fine one VCM is used. In this study, we propose the new structure of VCM for the fine actuator. It is 3 axis precision positioning stage for an aligner system. After we perform the optimal design of the stage to obtain the maximum force, which is related to the acceleration of the stage to accomplish throughput of product. And we controlled this fine stage with TDC. So we obtained 50nm resolution. So later more works will be done to obtain better accuracy.

• Smart Structures and Systems
• /
• 제18권5호
• /
• pp.1005-1028
• /
• 2016

Two novel semi-active control methods for a seismically excited nonlinear benchmark building equipped with magnetorheological dampers are presented and evaluated in this paper. While a primary controller is designed to estimate the optimal control force of a magnetorheological (MR) damper, the required voltage input for the damper to produce such desired control force is achieved using two different methods. The first technique uses an optimal compact Takagi-Sugeno-Kang (TSK) fuzzy inverse model of MR damper to predict the required voltage to actuate the MR dampers (TSKFInv). The other voltage regulator introduced here works based on the maximum and minimum capacities of MR damper at each time-step (MaxMin). Both semi-active algorithms developed here, use acceleration feedback only. The results demonstrate that both TSKFInv and MaxMin algorithms are quite effective in seismic response reduction for wide range of motions from moderate to severe seismic events, compared with the passive systems and performs better than original and Modified clipped optimal controller systems, known as COC and MCOC.

• Earthquakes and Structures
• /
• 제13권3호
• /
• pp.221-230
• /
• 2017

One of the important phases of probabilistic performance-based methodology is establishing appropriate probabilistic seismic demand models (PSDMs). These demand models relate ground motion intensity measures (IMs) to demand measures (DMs). The objective of this paper is selection of the optimal IMs in probabilistic seismic demand analysis (PSDA) of the RC high-rise buildings. In selection process features such as: efficiency, practically, proficiency and sufficiency are considered. RC high-rise buildings with core wall structural system are selected as a case study building class with the three characteristic heights: 20-storey, 30-storey and 40-storey. In order to determine the most optimal IMs, 720 nonlinear time-history analyses are conducted for 60 ground motion records with a wide range of magnitudes and distances to source, and for various soil types, thus taking into account uncertainties during ground motion selection. The non-linear 3D models of the case study buildings are constructed. A detailed regression analysis and statistical processing of results are performed and appropriate PSDMs for the RC high-rise building are derived. Analyzing a large number of results it are adopted conclusions on the optimality of individual ground motion IMs for the RC high-rise building.