• Proceedings of the IEEK Conference
• /
• 2002.06e
• /
• pp.63-66
• /
• 2002

This paper has been studied the parallel resonant inverter which controlling the constant power and tracking the load resonant frequency with PLL is possible, in order to minimize switching losses. The current-fed full-bridge type parallel resonant inverter of an induction heating system was composed of IGBT in switching device. For regulating the output power of an induction heating system, the Fuzzy controller is used. The Fuzzy controller makes the control signal for a stable power regulating control and when reference is changed, it is superior to adaptability. It has been evaluated a stable behavior for a noise with switching and a load disturbance.

• Journal of Power Electronics
• /
• v.12 no.6
• /
• pp.904-911
• /
• 2012

This paper presents a fuzzy adaptive speed controller that guarantees a fast dynamic behavior and a precise trajectory tracking capability for surfaced-mounted permanent magnet synchronous motors (SPMSMs). The proposed fuzzy adaptive control strategy is simple and easy to implement. In addition, the proposed speed controller is very robust to system parameter and load torque variations because it does not require any accurate parameter values. The global stability of the proposed control system is analytically verified. To evaluate the proposed fuzzy adaptive speed controller, both simulation and experimental results are shown under motor parameter and load torque variations on a prototype SPMSM drive system.

• Journal of Korea Multimedia Society
• /
• v.24 no.3
• /
• pp.345-354
• /
• 2021

Recently, deep learning-based image recognition systems have been adopted to various surveillance environments, but most of them are still picture-type object recognition methods, which are insufficient for the long term temporal analysis and high-dimensional situation management. Therefore, we propose a method recognizing the specific dangerous situation generated by human in real-time, and utilizing deep learning-based object analysis techniques. The proposed method uses deep learning-based object detection and tracking algorithms in order to recognize the situations such as 'trespassing', 'loitering', and so on. In addition, human's joint pose data are extracted and analyzed for the emergent awareness function such as 'falling down' to notify not only in the security but also in the emergency environmental utilizations.

• Fisheries and Aquatic Sciences
• /
• v.26 no.12
• /
• pp.698-707
• /
• 2023

Fish are an essential resource in human society, and while ecological research on them is challenging, it is absolutely necessary. Recent technologies enabled researchers to monitor underwater fish behavior. Acoustic signals, satellite-mediated location estimation, and light-based geolocation are powerful tools for tracking fish movements from freshwater to deep-sea habitats. These tools allow us to track various fish species and elucidate their ecology. Furthermore, based on these technologies, we can develop fisheries management plans and enhance aquaculture productivity. In this review, we also discuss challenges in improving current technologies and provide future recommendations for fish bio-logging studies.

• Journal of the Korean Geographical Society
• /
• v.47 no.3
• /
• pp.426-439
• /
• 2012

The goal of this study is to propose a new method to capture the qualitative person spatial behavior. Beyond tracking or indexing the change of the location of a person, the changes in the relationships between a person and its environment are considered as the main source for the formal model of this study. Specifically, this paper focuses on the movement behavior of a person near the boundary of a region. To capture the behavior of person near the boundary of regions, a new formal approach for integrating an object's scope of influence is described. Such an object, a spatio-temporally extended point (STEP), is considered here by addressing its scope of influence as potential events or interactions area in conjunction with its location. The formalism presented is based on a topological data model and introduces a 12-intersection model to represent the topological relations between a region and the STEP in 2-dimensional space. From the perspective of STEP concept, a prototype analysis results are provided by using GPS tracking data in real world.

• Journal of Environmental Science International
• /
• v.25 no.9
• /
• pp.1253-1267
• /
• 2016

After large-scale flooding damage occurred along the Imjin river in 1996, 1998, and 1999, the Hantan river flood control dam was planned, and it has since been under construction. Unlike existing dams in Korea, the Hantan river flood control dam will remain fully open except during high floods, when the dam will store flood water temporarily to reduce flood peaks and flood water volume downstream. During past flooding seasons, floating debris has caused difficulties in the management of large-scale dams. Most of the existing multipurpose dams in Korea have installed nets to collect floating debris based on many years of experience with and data about inflow and distribution of floating debris in the dams. For the Hantan river flood control dam, however, collection of data about inflow and distribution of floating debris is not possible as the dam is located near the border area between North and South Korea. In order to devise a preliminary plan to collect floating debris in the Hantan river flood control dam, an EFDC hydrodynamic model was used to analyze the behavior of floating debris during high floods. The Lagrangian particle tracking method was utilized to simulate the behavior of floating debris in the dam. Based on the analysis of paths and final destinations of the particles, seven collection points were selected where it seemed to be effective to collect floating debris, as debris is likely to accumulate there in high density.

• Korean Institute of Interior Design Journal
• /
• v.17 no.4
• /
• pp.84-91
• /
• 2008

The objective of this study is to investigate how visitors' behavior change in relation to the pattern of exhibition in art museum. For the purpose Busan Museum of Modern Art was selected as the sample in this study. And two patterns of exhibition, namely the planar exhibition and the three dimensional exhibition, were surveyed using the tracking movement method. The data gathered from random-sampling 60 visitors in a art museum was analyzed by the correlation analysis and T-test. The results are as follows; 1) There is no meaningful difference in visitors' behavior by the exhibition pattern in terms of their viewing area, viewing time, viewing distance, viewing velocity, viewing rate. 2) In contrast, there is meaningful difference in terms of the number of stops by visitors and the time spent without movement to see the objects; visitors stopped more often and stayed longer within the three dimensional exhibition space than the planar exhibition in a art museum. In sum, visitors extrinsically show no big behavioral difference between the planar exhibition and the three dimensional exhibition, but much difference in the way they communicate with objects; they show more active behavior within the three dimensional exhibition space.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.37 no.3
• /
• pp.259-266
• /
• 2013

An automated cell tracking system is used to automatically analyze and track the changes in cell behavior in time-lapse cell images acquired using a microscope with a cell culture. Clustering is the partial overlapping of neighboring cells in the process of cell change. Separating clusters into individual cells is very important for cell tracking. In this study, we proposed an algorithm for separating clusters by using ellipse fitting based on a direct least square method. We extracted the contours of clusters, divided them into line segments, and then produced their fitted ellipses using a direct least square method for each line segment. All of the fitted ellipses could be used to separate their corresponding clusters. In experiments, our algorithm separated clusters with average precisions of 91% for two overlapping cells, 84% for three overlapping cells, and about 73% for four overlapping cells.

• Journal of Ocean Engineering and Technology
• /
• v.23 no.2
• /
• pp.47-52
• /
• 2009

A Lagrangian particle tracking model coupled with the Princeton Ocean Model were used to estimate the average residence time of coastal water in Masan Bay, Korea. Our interest in quantifying the transport time scales in Masan Bay was stimulated by the search for a mechanistic understanding of this spatial variability, which is consistent with the concept of spatially variable transport time scales. Tidal simulation was calibrated through a comparison with the results of semi-diurnal current and water elevation measured at the tidal stations of Masan, Gadeokdo. In the model simulations, particles were released in eight cases, including slack before ebb, peak ebb, slack before flood, and peak flood, during both spring and neap tides. The averaged values obtained from the particle release simulations were used for the average residence times of the coastal water in Masan Bay. The average residence times for the southeastern parts of Somodo and the Samho River, Masan Bay were estimated to be about 20~50days and 70~80days, respectively. The spatial difference for the average residence time was controlled by the tidal currents and distance from the mouth of the bay. Our results might provide useful for understanding the transport and behavior of coastal water in a bay and might be used to estimate the dissimilative capacity for environmental assessment.

• Journal of Korean Society of Transportation
• /
• v.23 no.7 s.85
• /
• pp.137-148
• /
• 2005

The position information of individual vehicles on a road at every time instant can be used to analyze the microscopic behaviors of driving of each vehicle. The limited information obtained from previous imaging technology such as traffic volume and interval velocity cannot be used to explore such microscopic traffic conditions. Also, information gathering for the microscopic behaviors by manual analysis of captured video takes large amount of time and man-power. In the paper we develop the rule-based vehicle tracking technology from which the position information of individual vehicles on a road at every time instant can be automatically obtained. Also, we extract the position data of driving vehicles on a road, length of 130m for every 0.05 second, and calculate the velocity of each traced vehicles to compare with the real velocity for the verification of accuracy. In the future, this type of tracking techniques based on video analysis can be widely used to provide the practically important information of road traffic conditions and to analyze the academically important microscopic behaviors of driving patterns.