• Journal of Korean Institute of Industrial Engineers
• /
• v.28 no.2
• /
• pp.147-161
• /
• 2002

In novelty detection, one attempts to discriminate abnormal patterns from normal ones. Novelty detection is quite difficult since, unlike usual two class classification problems, only normal patterns are available for training. Auto-Associative Multi-Layer Perceptron (AAMLP) has been shown to provide a good performance based upon the property that novel patterns usually have larger auto-associative errors. In this paper, we give a mathematical analysis of 2-layer AAMLP's output characteristics and empirical results of 2-layer and 4-layer AAMLPs. Various activation functions such as linear, saturated linear and sigmoid are compared. The 2-layer AAMLPs cannot identify non-linear boundaries while the 4-layer ones can. When the data distribution is multi-modal, then an ensemble of AAMLPs, each of which is trained with pre-clustered data is required. This paper contributes to understanding of AAMLP networks and leads to practical recommendations regarding its use.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.24 no.4
• /
• pp.457-462
• /
• 2011

In this study, a simplified design procedure for friction dampers of a multi-story structure in order to reduce seismic response is proposed. To get insight for control effect of the structure with friction dampers is difficult, because of a nonlinear characteristic by a friction damper. Since a control force of a friction damper is influenced by coupling velocity between floors, adjoining modes are coupled. Thus structural response are derived by assuming steady-state response in resonance. As it is impossible that an exact solution is obtained for seismic load, first, a closed form solution can be achieved under harmonic vibration. Second, to convert a three-story building into a single-degree-of-freedom(SDOF) structure, modal analysis is performed. Third, an equivalent damping ratio is derived with utilizing closed form solution. And response reducing factor is proposed by it. Finally, friction force of a damper is designed for using response reducing factor, and then designed dampers are verified for seven seismic data. The nonlinear analysis results confirm the validity of the proposed procedure.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.8
• /
• pp.4769-4775
• /
• 2014

This paper discusses the dynamics stress of each boom, which occurs as a result of the conflicts on a multi-joint boom at the end effector and structure. In this process, CATIA was applied to create 3D modeling, ADAMS and ANSYS were then performed using mesh analysis by obtaining the stress data to create a MNF(Modal Neutral File) of multi-joint boom. Two types of MNF models were analyzed using ADAMS FFT(Fast Fourier Transform) performing to check the validity of each model. In this process, the models were verified by ADAMS, which performs the dynamic characteristics of conflict. The ADAMS Durability was used to analyze the maximum stress between the multi-joint boom and structure.

• Smart Structures and Systems
• /
• v.15 no.5
• /
• pp.1253-1270
• /
• 2015

In this paper Magnetic Charged System Search (MCSS) and Particle Swarm Optimization (PSO) are applied to the problem of damage detection using frequencies and mode shapes of the structures. The objective is to identify the location and extent of multi-damage in structures. Both natural frequencies and mode shapes are used to form the required objective function. To moderate the effect of noise on measured data, a penalty approach is applied. A variety of numerical examples including two beams and two trusses are considered. A comparison between the PSO and MCSS is conducted to show the efficiency of the MCSS in finding the global optimum. The results show that the present methodology can reliably identify damage scenarios using noisy measurements and incomplete data.

• Earthquakes and Structures
• /
• v.10 no.5
• /
• pp.1049-1066
• /
• 2016

An approximate analysis is presented for multi-story setback buildings subjected to ground motions. Setback buildings with mass and stiffness discontinuities are common in modern architecture and quite often they are asymmetric in plan. The proposed analysis provides basic dynamic data (frequencies and peak values of base resultant forces) and furthermore an overview of the building response during a ground excitation. The method is based on the concept of the equivalent single story system, which has been introduced by the author in earlier papers for assessing the response of uniform in height buildings. As basic quantities of the dynamic response of elastic setback buildings can be derived by analyzing simple systems, a structural layout of minimum elastic rotational response can be easily constructed. The behavior of such structural configurations, which is basically translational into the elastic phase, is also examined into the post elastic phase when the strength assignment of the various bents is based on a planar static analysis under a set of lateral forces simulating an equivalent 'seismic loading'. It is demonstrated that the almost concurrent yielding of all resisting elements preserves the translational response, attained at the end of the elastic phase, to the post elastic one.

• Wind and Structures
• /
• v.21 no.5
• /
• pp.537-564
• /
• 2015

Excessive motions in buildings cause occupants to become uncomfortable and nervous. This is particularly detrimental to the tenants and ultimately the owner of the building, with respect to financial considerations. Serviceability issues, such as excessive accelerations and inter-story drifts, are more prevalent today due to advancements in the structural systems, strength of materials, and design practices. These factors allow buildings to be taller, lighter, and more flexible, thereby exacerbating the impact of dynamic responses. There is a growing need for innovative and effective techniques to reduce the serviceability responses of these tall buildings. The current study considers a case study of a real building to show the effectiveness and robustness of the TLD in reducing the coupled lateral-torsional motion of this high-rise building under wind loading. Three unique multi-modal TLD systems are designed specifically to mitigate the torsional response of the building. A procedure is developed to analyze a structure-TLD system using High Frequency Force Balance (HFFB) test data from the Boundary Layer Wind Tunnel Laboratory (BLWTL) at the University of Western Ontario. The effectiveness of the unique TLD systems is investigated. In addition, a parametric study is conducted to determine the robustness of the systems in reducing the serviceability responses. Three practical parameters are varied to investigate the robustness of the TLD system: the height of water inside the tanks, the amplitude modification factor, and the structural modal frequencies.

• Journal of the Korea Convergence Society
• /
• v.9 no.1
• /
• pp.419-431
• /
• 2018

In this paper, we propose a novel method for monitoring mood trend of Twitter users by analyzing their daily tweets for a long period. Then, to more accurately understand their tweets, we analyze all types of content in tweets, i.e., texts and emoticons, and images, thus develop a multimodal sentiment analysis method. In the proposed method, two single-modal analyses first are performed to extract the users' moods hidden in texts and images: a lexicon-based and learning-based text classifier and a learning-based image classifier. Thereafter, the extracted moods from the respective analyses are combined into a tweet mood and aggregated a daily mood. As a result, the proposed method generates a user daily mood flow graph, which allows us for monitoring the mood trend of users more intuitively. For evaluation, we perform two sets of experiment. First, we collect the data sets of 40,447 data. We evaluate our method via comparing the state-of-the-art techniques. In our experiments, we demonstrate that the proposed multimodal analysis method outperforms other baselines and our own methods using text-based tweets or images only. Furthermore, to evaluate the potential of the proposed method in monitoring users' mood trend, we tested the proposed method with 40 depressive users and 40 normal users. It proves that the proposed method can be effectively used in finding depressed users.

• Journal of Korea Multimedia Society
• /
• v.23 no.10
• /
• pp.1307-1317
• /
• 2020

In recent years, the PV plant is gradually increasing in korea and has begun to focus on PV system O&M(Operation and Management). The legacy PV system can hardly be gathering realtime information. That is why it is difficult for the facility to respond immediately to situations in which problems arise. This leads to economic losses due to reduced power generation. In this study, we implemented an RMC(Remote Maintenance Control) gateway that supports multi-area equipment(inverters, junction boxes, switchboards, environmental sensors) that are not limited to specific manufacturers. It provides a function to visualize the collected data. Users can easily check the power generation and operation status of PV system equipment. In the future, we will develop an algorithm capable of analyzing beta factors that cause equipment performance degradation, parts aging, and deterioration using data collected from facilities. We intend to use this algorithm as a fundamental technology for O&M by transplanting it to the RMC gateway.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.40 no.2
• /
• pp.91-108
• /
• 2022

In most cases, optical images have been used as training data of DL (Deep Learning) models for object detection, recognition, identification, classification, semantic segmentation, and instance segmentation. However, properties of 3D objects in the real-world could not be fully explored with 2D images. One of the major sources of the 3D geospatial information is DSM (Digital Surface Model). In this matter, characteristic information derived from DSM would be effective to analyze 3D terrain features. Especially, man-made objects such as buildings having geometrically unique shape could be described by geometric elements that are obtained from 3D geospatial data. The background and motivation of this paper were drawn from concept of the intrinsic image that is involved in high-level visual information processing. This paper aims to extract buildings after classifying terrain features by training DL model with DSM-derived information including slope, aspect, and SRI (Shaded Relief Image). The experiments were carried out using DSM and label dataset provided by ISPRS (International Society for Photogrammetry and Remote Sensing) for CNN-based SegNet model. In particular, experiments focus on combining multi-source information to improve training performance and synergistic effect of the DL model. The results demonstrate that buildings were effectively classified and extracted by the proposed approach.

• Journal of Korea Multimedia Society
• /
• v.12 no.5
• /
• pp.708-716
• /
• 2009

Applications which are implemented on Table Top Display are controlled by hands, so that they support an intuitive interface to users. Users feel the real sense when they interact on the virtual scene in Table Top Display. However, most of conventional augmented reality applications on Table Top Display satisfy only visual sense. In this paper, we propose an interface that supports multi-modal sense in that tactile sense is utilized for augment reality by vibrating a physical control unit when it collides to virtual objects. Users can feel the collision in addition to visual scene. The proposed system facilitates tactile augmented reality through an air hockey game. A physical control unit vibrates when it receives virtual collision data over wireless communication. Since the use of tabletop display environment is extended with a tactile sense based physical unit other than hand, it provides a more intuitive interface.