• Journal of Sensor Science and Technology
• /
• v.5 no.6
• /
• pp.35-42
• /
• 1996

In this paper, a development of model based battery SOC indicator is described. The proposed method is independent upon initial SOC, is reliable on the sudden change of load, and could estimate the available driving distance. The mathematical model of battery which has relation of the current, voltage and SOC estimates the SOC by least square estimation to minimize the error between measured voltage and estimated voltage. For experiment, the charging and discharging system using computer was designed to acquire the current and voltage data for model. The feasibility in electric vehicle was confirmed by variable load testing using the developed SOC indicator by stand-alone type microcontroller.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.27 no.6
• /
• pp.1431-1439
• /
• 2017

In the security field, log analysis is important to detect malware or abnormal behavior. Recently, image visualization techniques for malware dectection becomes to a major part of security. These techniques can also be used in online games. Users can leave a game when they felt bad experience from game bot, automatic hunting programs, malicious code, etc. This churning can damage online game's profit and longevity of service if game operators cannot detect this kind of events in time. In this paper, we propose a new technique of PNG image conversion based churn prediction to improve the efficiency of data analysis for the first. By using this log compression technique, we can reduce the size of log files by 52,849 times smaller and increase the analysis speed without features analysis. Second, we apply data mining technique to predict user's churn with a real dataset from Blade & Soul developed by NCSoft. As a result, we can identify potential churners with a high accuracy of 97%.

• Tunnel and Underground Space
• /
• v.12 no.2
• /
• pp.107-114
• /
• 2002

It is the common practice to over design the reinforcement for the secondary tunnel lining due to the lack of rational insight into the ground loosening loads. and due to the conservative application of the empirical design methods. The main loads of the secondary lining are the ground Loosening loads and the ground water pressure, and the ground load is critical in the reinforcement design of the secondary lining in the case of drained tunnel. If the external load is absent around a tunnel, the reasons of the load far secondary tunnel lining are the deterioration of the primary supports such as shotcrete, steel rib, and rockbolts. Accordingly, the analysis method considering the ground-primary supports-secondary lining interaction should be required tar the rational design of the secondary tunnel lining. In this paper, the interaction was conceptually described by the simple mass-spring model and the load transfer from the ground and primary supports to the secondary lining is showed by the ground-primary supports-secondary lining reaction curves fur the theoretical solution of a circular tunnel. And also, the application of this proposed model to numerical analysis is verified in order to check the potential far the tunnel with the complex analysis conditions.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.3B
• /
• pp.311-317
• /
• 2009

The traditional network anomaly detection systems execute the threshold-based detection without considering dynamic network environments, which causes false positive and limits an effective resource utilization. To overcome the drawbacks, we present the adaptive network anomaly detection model based on artificial immune system (AIS) in centralized network. AIS is inspired from human immune system that has learning, adaptation and memory. In our proposed model, the interaction between dendritic cell and T-cell of human immune system is adopted. We design the main components, such as central node and router node, and define functions of them. The central node analyzes the anomaly information received from the related router nodes, decides response policy and sends the policy to corresponding nodes. The router node consists of detector module and responder module. The detector module perceives the anomaly depending on learning data and the responder module settles the anomaly according to the policy received from central node. Finally we evaluate the possibility of the proposed detection model through simulation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.10
• /
• pp.1737-1743
• /
• 2008

A weather radar extracts the weather information from the return echoes which consist of scattered electromagnetic wave signals from rain, cloud and dust particles, etc. The characteristics of Doppler weather signal and ground clutter should be analyzed to extract the accurate weather information. However, the conventional symmetric weather Doppler model is somewhat inappropriate in representing various weather situations. Therefore, the improved model is suggested to describe the skewness in the Doppler spectrum model. Using the suggested model, many various weather signals can be simulated efficiently in time and spectral domain according to weather situations, operation environment and system characteristics. This simulation method may be very helpful in verifying the accuracy of the weather information extraction algorithms and developing the new system for further performance improvement.

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.184-196
• /
• 2021

This paper proposes statistics adaptive linear regression-based object size prediction method for object detection. YOLOv2 and YOLOv3, which are typical deep learning-based object detection algorithms, designed the last layer of a network using statistics adaptive exponential regression model to predict the size of objects. However, an exponential regression model can propagate a high derivative of a loss function into all parameters in a network because of the property of an exponential function. We propose statistics adaptive linear regression layer to ease the gradient exploding problem of the exponential regression model. The proposed statistics adaptive linear regression model is used in the last layer of the network to predict the size of objects with statistics estimated from training dataset. We newly designed the network based on the YOLOv3tiny and it shows the higher performance compared to YOLOv3 tiny on the UFPR-ALPR dataset.

• Journal of The Korean Society For Urban Railway
• /
• v.6 no.4
• /
• pp.363-372
• /
• 2018

Simulator is used to verifying the function and performance of train control system before verifying with actual train. In this case, it is important that the simulation result should be coincide with the result with actual train. In this paper, the process of the development of automatic train operation (ATO) is described. ATO is in charge of automatic train control such as speed regulation and precision stop control. Identical interfaces from the ATO to the actual train was made in the simulator. Therefore ATO communicates to the simulator in the same way to the actual train. Futhermore, the train dynamic properties was measured by experiments and these were applied to the train model. Hence the response of the train in the simulator to the acceleration command is similar to that of the actual train. The simulation result of precision stop control is compared with the result in the actual train test to show the fidelity of the train model derived in the study and the superiority of this simulator.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.47 no.6
• /
• pp.438-445
• /
• 2019

In this paper, we present a new approach to extract the g-sensitivity scale-factor error for a MEMS gyroscope. MEMS gyroscopes, based on the use of both angular momentum and the Coriolis effect, have a g-sensitivity error due to mass unbalance. Generally, the g-sensitivity error is not considered in general use of gyroscopes, but it deserves our attention if we are to develop for tactical class performance and reliability. The g-sensitivity error during vehicle flight increases navigation error; so it must be analyzed and compensated for the use of MEMS IMU for high dynamics vehicle systems. Therefore, we analyzed how to extract the g-sensitivity scale-factor error from the inertial sensor error model. Furthermore we propose a new method to extract the g-sensitivity error using flight motion simulator. We verified our proposed method with experimental results.

• Journal of the Korean Geotechnical Society
• /
• v.36 no.12
• /
• pp.125-132
• /
• 2020

Embedded joints in the rock mass are a major constituent influencing its mechanical behavior. Numerical analysis requires a rigorous modeling methodology for the rock mass with detailed information regarding joint properties, orientation, spacing, and persistence. This paper provides a mechanical model for a jointed rock mass based on the implicit joint-continuum approach. Stiffness tensors for rock mass are evaluated for an assemblage of intact rock separated by sets of joint planes. It is a linear summation of compliance of each joint sets and intact rock in the serial stiffness system. In the application example, kinematic analysis for a planar failure of rock slope is comparable with empirical daylight envelope and its lateral limits. Since the developed implicit joint-continuity model is formulated on a continuum basis, it will be a major tool for the numerical simulations adopting published plenteous thermal-hydro-chemical experimental results.

• The Journal of Society for e-Business Studies
• /
• v.26 no.1
• /
• pp.29-41
• /
• 2021

Recently, there are increasing attempts to utilize deep learning methodology in the fashion industry. Accordingly, research dealing with various fashion-related problems have been proposed, and superior performances have been achieved. However, the studies for fashion style classification have not reflected the characteristics of the fashion style that one outfit can include multiple styles simultaneously. Therefore, we aim to solve the multi-label classification problem by utilizing the dependencies between the styles. A multi-label recognition model based on a graph convolution network is applied to detect and explore fashion styles' dependencies. Furthermore, we accelerate model training and improve the model's performance through transfer learning. The proposed model was verified by a dataset collected from social network services and outperformed baselines.