• Journal of KIISE:Software and Applications
• /
• v.36 no.4
• /
• pp.273-282
• /
• 2009

Missing data is one of the common problems in building analysis or prediction models using software project data. Missing imputation methods are known to be more effective missing data handling method than deleting methods in small software project data. While K nearest neighbor imputation is a proper missing imputation method in the software project data, it cannot use non-missing information of incomplete project instances. In this paper, we propose an approach to missing data imputation for numerical software project data by combining K nearest neighbor and maximum likelihood estimation; we also extend the average absolute error measure by normalization for accurate evaluation. Our approach overcomes the limitation of K nearest neighbor imputation and outperforms on our real data sets.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2000.11a
• /
• pp.316-321
• /
• 2000

Die Casting is one of the forming methods to manufacture large number of products with short period time and clean surface by high forming pressure and temperature of cast alloy. Die design is composed of selection of cast alloy, design of die casting product, runner and gate design etc. In reality, however, die design of die casting has been performed by trial and error method, which cause economic and time loss. This paper describes a research work of developing computer-aided design of die casting product and die design. Approach to the CAD system has been written in AutoLISP on the AutoCAD with personal computer. In this study, die design system for gate of die casting process has been developed to present algorithm for automation of die design, especially runner-gate system. As forming process and die design system using 3-D geometry handling are integrated with technology of process planning, die design is possible to set. In addition, specific rules and equations for the runner-gate system have been presented to avoid too many trails and errors with expensive equipment. It is possible for engineers to make automatic and efficient die design of die casting and it will result in reduction of expense and time to be required. An example is applied to cap-shaped casting using proposed algorithm.

• Spatial Information Research
• /
• v.13 no.4 s.35
• /
• pp.355-364
• /
• 2005

Vector data sets (e.g. maps) are currently major sources of displaying, querying, and identifying locations of spatial features in a variety of applications. Especially in mobile environment, the needs for using spatial data is increasing, and the relative large size of vector maps need to be smaller. Recently, there have been several studies about vector map compression. There was clustering-based compression method with novel encoding/decoding scheme. However, precedent studies did not consider that spatial data have to be updated periodically. This paper explores the problem of existing clustering-based compression method. We propose an adaptive approximation method that is capable of handling data updates as well as reducing error levels. Experimental evaluation showed that when an updated event occurred the proposed adaptive approximation method showed enhanced positional accuracy compared with simple cluster based compression method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.3
• /
• pp.356-363
• /
• 2014

This paper covers improvement of the humanoid robot platform HUBO2, known as the HUBO2+. As a necessity of the growth of the humanoid platform, a robust, reliable and user friendly platform is needed. From this standpoint, HUBO2+ is the most improved humanoid robot platform in the HUBO series. The mechanical design has been changed to increase the movable range and to stop joint compulsion. Additionally, all of the electrical parts are re-designed to be un-breakable in an unexpected situation. A smart power controller with robot status check panel is attached on the back. Additionally, a diagnosis tool, the HUBO-i, has been developed. Moreover, each joint motor controller of HUBO2+ has a Protection Function and a PODO system is provided for handling the robot easily.

• Journal of Mechanical Science and Technology
• /
• v.20 no.5
• /
• pp.658-667
• /
• 2006

Fractional derivative models, which are used to describe the viscoelastic behavior of material, have received considerable attention. Thus it is necessary to put forward the analysis solutions of dynamic systems containing a fractional derivative. Although previously reported such kind of fractional calculus-based constitutive models, it only handles the particularity of rational number in part, has great limitation by reason of only handling with particular rational number field. Simultaneously, the former study has great unreliability by reason of using the complementary error function which can't ensure uniform real number. In this paper, a new approach is proposed for an analytical scheme for dynamic system of a spring-mass-damper system of single-degree of freedom under general forcing conditions, whose damping is described by a fractional derivative of the order $0<{\alpha}<1$ which can be both irrational number and rational number. The new approach combines the fractional Green's function and Laplace transform of fractional derivative. Analytical examples of dynamic system under general forcing conditions obtained by means of this approach verify the feasibility very well with much higher reliability and universality.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.580-583
• /
• 2003

This paper presents the motion detection algorithm that can run robustly about recusive motion. The existing motion detection algorithm that uses difference image is robustly in some degree brightness or noise, but it frequently causes false alarms to temporal clutter, at the repetitive motion within a certain area. We developed a motion detection algorithm using mean absoulte error(MAE) which calculates the set of Moving regions and performs block matching. The experimental results revealed that our approach is superior to existing methodologies to handling various temporal clutter.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.27 no.4
• /
• pp.137-142
• /
• 1999

The purpose of this study is to develope a CAD-based tool for rasterization of polygonal vector map in AutoCAD. To identity the layer property of polygonal entity with user-defined coordinates as topology, algorithm in processing entity data of selection set that intersected with scan line was used, and the layers were extracted sequentially by sorted intersecting points in data-list. In addition to the functions for querying and modifying topology, two options for mapping were set up to construct plan projection type and to change meshes' properties in existing DTM data. In case of plan projection type, user-defined cell size of 3DFACE mesh is available for more detailed edge, and topological draping on landform can be executed in case of referring DTM data as an AutoCAD's drawing. The concept of algorithm was simple and clear, but some unexpectable errors were found in detecting intersected coordinates that were AutoCAD's error, not the utility's. Also, the routines to check these errors were included in algorithmic processing. Developed utility named MESHMAP was written in entity data control functions of AutoLISP language and dialog control language(DCL) for the purpose of user-oriented interactive usage. MESHMAP was proved to be more effective in data handling and time comparing with GRIDMAP module in LANDCADD which has similar function.

• Journal of Information Processing Systems
• /
• v.12 no.3
• /
• pp.511-524
• /
• 2016

The Virtual Local Area Network (VLAN) has been used for a long time in campus and enterprise networks as the most popular network virtualization solution. Due to the benefits and advantages achieved by using VLAN, network operators and administrators have been using it for constructing their networks up until now and have even extended it to manage the networking in a cloud computing system. However, their configuration is a complex, tedious, time-consuming, and error-prone process. Since Software Defined Networking (SDN) features the centralized network management and network programmability, it is a promising solution for handling the aforementioned challenges in VLAN management. In this paper, we first introduce a new architecture for campus and enterprise networks by leveraging SDN and OpenFlow. Next, we have designed and implemented an application for easily managing and flexibly troubleshooting the VLANs in this architecture. This application supports both static VLAN and dynamic VLAN configurations. In addition, we discuss the hybrid-mode operation where the packet processing is involved by both the OpenFlow control plane and the traditional control plane. By deploying a real test-bed prototype, we illustrate how our system works and then evaluate the network latency in dynamic VLAN operation.

• Korean System Dynamics Review
• /
• v.3 no.2
• /
• pp.49-68
• /
• 2002

The intent of this study is to develop system dynamics model for assessment of organizational and human factors in nuclear power plant which can contribute to secure the nuclear safety. Previous studies are classified into two major approaches. One is engineering approach such as ergonomics and probability safety assessment(PSA). The other is social science approach such like sociology, organization theory and psychology. Both have contributed to find organization and human factors and to present guideline to lessen human error in NPP. But, since these methodologies assume that relationship among factors is independent they don't explain the interactions among factors or variables in NPP. To overcome these limits, we have developed system dynamics model which can show cause and effect among factors and quantify organizational and human factors. The model we developed is composed of 16 functions of job process in nuclear power, and shows interactions among various factors which affects employees' productivity and job quality. Handling variables such like degree of leadership, adjustment of number of employee, and workload in each department, users can simulate various situations in nuclear power plant in the organization side. Through simulation, user can get insight to improve safety in plants and to find managerial tools in the organization and human side. Analyzing pattern of variables, users can get knowledge of their organization structure, and understand stands of other departments or employees. Ultimately they can build learning organization to secure optimal safety in nuclear power plant.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.372-372
• /
• 2000

In this paper, we propose Neuro Fuzzy Polynomial Networks(NFPN) based on Polynomial Neural Network(PNN) and Neuro-Fuzzy(NF) for model identification of complex and nonlinear systems. The proposed NFPN is generated from the mutually combined structure of both NF and PNN. The one and the other are considered as the premise part and consequence part of NFPN structure respectively. As the premise part of NFPN, NF uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. As the consequence part of NFPN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. NFPN is available effectively for multi-input variables and high-order polynomial according to the combination of NF with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. In order to evaluate the performance of proposed models, we use the nonlinear function. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously.