• Journal of Environmental Science International
• /
• v.5 no.1
• /
• pp.83-91
• /
• 1996

A recycling soap was prepared from non-cooking oils. The effects of physlcal and chemical properties of the recycling soap on biodegradation are expected to be different due to the thermal histories of the non-cooking oils. Therefore, the biodegradation rate of the recycling soap was studied by using Klebssella Pneumoniae(K. pneumoniae), and the growth rate of K. pnewoniae in soap solution was observed. The biodegradation rate of the recycling soap appeared to be slower as the thermal histories of the non-cooking oils became larger. This might be resulted from hydrolysis, in which the ester bonds in the oils are broken to produce hydroxyl group. It was also observed that the growth rate of the microorganism decreased with the increase in the thermal histories of the oils. As a result, it is desired that recycling soap should be produced from the non-cooking oils with the prober ranges of thermal histories to reduce water contamination. The non-cooking oils with larger thermal histories are considered to be recycling through the cracking process before used. Key Words : non-cooking oils, recycling soap, thermal history, biodegradation, microorganism growth.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.2
• /
• pp.120-125
• /
• 2014

Conventional handheld metal detectors use a single induction coil to detect the metallic parts of explosive objects, and the detector generates an acoustic signal from its magnetic response to a metallic object so that an operator can confirm the existence of mines. Though metal detectors have very useful detection mechanisms to find mines, it is easy to cause a high false alarm ratio due to the detection of non-explosive metallic items such as cans, nails and other pieces of metal, etc. Also, because of the physical characteristic of a metal detector it is hard to detect non-metallic objects such as mines made of wood or plastic. Furthermore, the operator must move it to the left and right slowly and repeatedly to attain enough sensor signals to confirm the existence of mines using only a monotonous acoustic signal. To resolve the disadvantages of handheld detectors, many new approaches have been attempted, such as an arrayed detector and a visualization algorithm based on metal/non-metal sensor. In this paper, we introduce a visualization algorithm with a metal/non-metal complex sensor, an arrayed metal/non-metal sensor and the their testing and evaluation.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.1-11
• /
• 2023

Federated learning (FL) is a ground breaking machine learning paradigm that allow smultiple participants to collaboratively train models in a cloud environment, all while maintaining the privacy of their raw data. This approach is in valuable in applications involving sensitive or geographically distributed data. However, one of the challenges in FL is dealing with heterogeneous and non-independent and identically distributed (non-IID) data across participants, which can result in suboptimal model performance compared to traditionalmachine learning methods. To tackle this, we introduce FedGCD, a novel FL algorithm that employs Graph Neural Network (GNN)-based community detection to enhance model convergence in federated settings. In our experiments, FedGCD consistently outperformed existing FL algorithms in various scenarios: for instance, in a non-IID environment, it achieved an accuracy of 0.9113, a precision of 0.8798,and an F1-Score of 0.8972. In a semi-IID setting, it demonstrated the highest accuracy at 0.9315 and an impressive F1-Score of 0.9312. We also introduce a new metric, nonIIDness, to quantitatively measure the degree of data heterogeneity. Our results indicate that FedGCD not only addresses the challenges of data heterogeneity and non-IIDness but also sets new benchmarks for FL algorithms. The community detection approach adopted in FedGCD has broader implications, suggesting that it could be adapted for other distributed machine learning scenarios, thereby improving model performance and convergence across a range of applications.

• International Journal of Advanced Culture Technology
• /
• v.8 no.1
• /
• pp.243-247
• /
• 2020

In the era of the fourth industrial revolution, safety of disasters is being emphasized above all else, and electric fires are most frequent during disasters, and human and property damage is very serious. In this paper, we propose a study that can determine the efficiency of evacuation and non-evacuation guidance due to the large difference in casualties depending on the traffic line in the case of fire. Evacuation guidance was assumed to be a situation in which adequate evacuation routes were guided by a recorded voice or a trained staff, and non-evacuation guidance was assumed to be a situation without anything. Evacuation simulations were carried out using a evacuation simulation tool called PATHFINDER and SIMULEX for the analysis of the efficiency of evacuation and non-evacuation guidance. As a result, the evacuation time was similar, but in the case of non-evacuation guidance, it was not guided to the safe zone, which could cause serious damage.

• Journal of ILASS-Korea
• /
• v.14 no.2
• /
• pp.49-58
• /
• 2009

Non-edible vegetable oils instead of edible vegetable oils as a substitute for diesel fuel are getting a renewed attention because of global reduction of green house gases and concerns for long-term food and energy security. Out of various non-edible vegetable oils, karanja, mahua, linseed, rubber seed and cotton seed oils are selected in this study. A brief review of recent works related to the application of the above five vegetable oils and its derivatives in CI engines is presented. The production technologies of biodiesel based on non-edible vegetable oils are introduced. Problems in vegetable oil or biodiesel fuelled CI engine are included. In addition, future works related to spray characteristics of non-edible vegetable oil or biodiesel from it are discussed. The biodiesel fuel, irrespective of the feedstock used, results in a decrease in the emission of hydrocardon (HC), carbon monoxide (CO), particulate matter (PM) and sulphur dioxide ($SO_2$). It is also said to be carbon neutral as it contributes no net carbon dioxide to the atmosphere. Only oxides of nitrogen (NOx) are reported to increase which is due to oxygen content in the biodiesel fuel. The systematic assessment of spray char-acteristics of neat vegetable oils and its blends, neat biodiesel and its blends f3r use as diesel engine fuels is required.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1999.10a
• /
• pp.403-410
• /
• 1999

General1y, H-section is used for columns and beams in the middle and low steel building, But it has a strong and weak axis. Thus if H-section is used for columns, the structure needs reinforcement on the weak axis. Therefore recently, square holler section(S.H.S) is used for columns because it is able to coiler the vulnerability of H-section. Structural analysis is usually executed under the assumption that connections are either ideally pinned joint or fully rigid joint. Actually all connections are semi-rigid which possess a rotational stiffness. Therefore it can be designed economically as using the property of connections which has a rotational stiffness. This paper presents a prediction model curve which is fitted Kishi-Chen power Model about the behavior of connection between H-beam and S.H.S column. Non-linear analysis program was considered the non-linearity of semi-rigid connection and the geometrical non-linearity under the effect of axial force. It was programed by FORTRAN90 and Visual Basic.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2006.11a
• /
• pp.241-244
• /
• 2006

It is essential to develop a large capacity, non-contact nondestructive inspection system having high reliability to investigate repaired and strengthened structures. Nowadays, an infrared camera is widely used in non-contact nondestructive inspection system. Because an infrared camera is sensitive to the surrounding environment, it is necessary to improve a sensitivity of thermal image information and a relationship between defects and thermal image information. In this papaer, presented is an optimum void detection chart for the optimum conditions to detect infrared rays from inside and outside defects like voids and cracks in concrete structures using extensive computer simulation. Sensitivity studies are performed with respect to variables influencing the temperature distribution such as heating temperature, heating time, and geometries of defect, etc. It may be stated that it could be successfully utilized for the non-contact nondestructive inspection system to detect defects in concrete structures.

• Journal of Advanced Marine Engineering and Technology
• /
• v.27 no.5
• /
• pp.641-647
• /
• 2003

The use of infrared range-finder sensor as the environment recognition system for mobile robot have the advantage of low sensing cost compared with the use of other vision sensor such as laser finder CCD camera. However, it is not easy to find the previous works on the use of infrared range-finder sensor for a mobile robot because of the non-linear characteristic of that. This paper describes the error due to non-linearity of a sensor and the correction of it using neural network. The neural network consists of multi-layer perception and Levenberg-Marquardt algorithm is applied to learning it. The effectiveness of the proposed algorithm is verified from experiment.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.22 no.8
• /
• pp.1073-1082
• /
• 1998

When thermal diffusivity is measured by laser flash method, the thermal diffusivity call be calculated front the assumption of the uniformly heated whole surface of the specimen. It has been known that the approximate 5% error is made by the non-uniform energy distribution on the specimen surface of laser pulse heat source. In this study, to obtain the highly-uniformed laser beam, which has both the low non-uniform heating error from non-uniform laser beam and the energy loss, research was carried out on no transmitting loss by optical fiber and high repetitions. In addition, heating error and thermal diffusivity were measured as the measuring positions were varied and compared with the results using the uniform and the non-uniform laser beams. In addition, dole to using the uniformalized laser beam, the whole surface of the specimen was heated uniformly and as a result, it was the thought that this was very effective to reduce the variations of the errors of the thermal diffusivity as the measuring positions were varied. It can be obtained that when the thermal diffusivity of POCO-AXM-5Q1 of SRM in NBS was measured with both the uniform and the non-uniform laser beams, the dispersion error of the former was from 2 to 2.5%, which was more improved than that of the latter.

• International Journal of Contents
• /
• v.5 no.3
• /
• pp.8-13
• /
• 2009

The recent Web service field emerges as the fastest growing IT paradigm as a result of the increasing interest in SOA (Services-Oriented Architecture) and the expansion of B2B market. With an increasing number of Web service that provide similar features, it becomes more important to provide the most appropriate service for the user's request. A service user in general requires the quality information of Web service when selecting a service among a number of similar Web services. Yet, finding a service through UDDI (Universal Description, Discovery and Integration) does not consider the non-functional aspects of users because it is only based on the functional aspects. That is, the quality, non-functional aspect will be an important factor for the mutual success of the user and provider. Using 3 factors in the Qos factors of the existing studies: the execution cost, reliability, and the quality level, the QoS of Web service is saved and the factors for the QoS are recorded in order to consider non-functional factors when selecting a Web service in this study. The Quality Broker determines the rank and shows the desired result of the service for users. The Quality Broker suggested in this thesis can be used to select a Web Service that considers the user-oriented and non-functional factors.