• Journal of Broadcast Engineering
• /
• v.16 no.6
• /
• pp.1009-1017
• /
• 2011

Fast generation of digital hologram is of importance for real-time holography broadcasting. In this paper, we propose such a method that parallelizes the Computer-Generated Holography (CGH) algorithm for digital hologram generation and make it faster using Multi Graphic Processing Unit (Multi-GPU) with help of the Compute Unified Device Architecture (CUDA) and the Open Multi-Processing (OpenMP). In addition, we propose optimization methods such as fixation variable, vectorization, and loop unrolling for making the CGH algorithm much faster. Experimental results show that our method is about 9,700 times faster than a CPU-based one.

• Journal of Advanced Marine Engineering and Technology
• /
• v.34 no.6
• /
• pp.858-863
• /
• 2010

In order to enhance encoding speed in dynamic image encoding using H.264/AVC, reducing the time for motion estimation which takes a large portion of the processing time is very important. An approach using graphics processing unit(GPU) as a coprocessor to assist the central processing unit(CPU) in computing massive data, will be a way to reduce the processing time. In this paper, we present an efficient block-level parallel algorithm for the motion estimation(ME) on a computer unified device architecture(CUDA) platform developed in general-purpose computation on GPU. Experiments are carried out to verify the effectiveness of the proposed algorithm.

• Journal of the Korean Society for Railway
• /
• v.17 no.2
• /
• pp.140-147
• /
• 2014

Line planning is an important step to determine the optimal frequencies of trains given the forecasted demand for each train type. The main input data for line planning is the leg traffic demand which can be derived using suitable traffic assignment models. However most assignment models require a line plan, in other words, train frequencies or headways, so that inconsistent results just by the procedural approach to find an optimal line plan after determining leg traffic can be avoided. This paper suggests a unified model that can consider the traffic assignment and line planning, simultaneously. We further provide an elaborated approximation algorithm and, finally, provide experimental results determined for the Korean railway network.

• The Journal of Korea Robotics Society
• /
• v.13 no.2
• /
• pp.129-141
• /
• 2018

This paper proposes a unified framework that overcomes four motion constraints including joint limit, kinematic singularity, algorithmic singularity and obstacles. The proposed framework is based on our previous works which can insert or remove tasks continuously using activation parameters and be applied to avoid joint limit and singularity. Additionally, we develop a method for avoiding obstacles and combine it into the framework to consider four motion constraints simultaneously. The performance of the proposed framework was demonstrated by simulation tests with considering four motion constraints. Results of the simulations verified the framework's effectiveness near joint limit, kinematic singularity, algorithmic singularity and obstacles. We also analyzed sensitivity of our algorithm near singularity when using closed loop inverse kinematics depending on magnitude of gain matrix.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.65 no.5
• /
• pp.811-818
• /
• 2016

This paper suggests a structure of power control system in floating wave-offshore wind hybrid power generation system. We have developed an unified SCADA(Supervisory Control and Data Acquisition) system which can be used to monitor and control PCS(Power Conversion System) based on IEC61850. The SCADA system is essential to perform the algorithm like proportional distribution and data acquisition, monitoring, active power, reactive power control in hybrid power generation system. IEC61850 is an international standard for electrical substation automation systems. It was made to compensate the limitations of the legacy industrial protocols such as Modbus. In order to test the proposed SCADA system and algorithm, we have developed the wind-wave simulator based Modbus. We have designed a protocol conversion device based on real-time Linux for the communication between Modbus and IEC61850. In this study, SCADA system consists of four 3MW class wind turbines and twenty-four 100kW class wave force generator.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.50 no.8
• /
• pp.388-394
• /
• 2001

This paper presents how to determine the optimal operating points of Unified Power Flow controllers (UPFC) the line flow control of which can enhance system security level. In order to analyze the effect of these devices on the power system, the decoupled model has been employed as a mathematical model of UPFC for power flow analysis. The security index that indicates the level of congestion of transmission line has been proposed and minimized by iterative method. The sensitivity of objective function for control variables of and UPFC has been derived, and it represents the change in the security index for a given set of changes in real power outputs of UPFC. The proposed algorithm with sensitivity analysis gives the optimal set of operating points of multiple UPECs that reduces the index or increases the security margin and Marquart method has been adopted as an optimization method because of stable convergence. The algorithm is verified by the 10-unit 39-bus New England system that includes multiple FACTS devices. The simulation results show that the power flow congestion can be relieved in normal state and the security margin can be guaranteed even in a fault condition by the cooperative operation of multiple UPECs.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.1916-1917
• /
• 2011

A development of a new moving obstacle avoidance algorithm using a delay-time Compensation for a network-based autonomous mobile robot is proposed in this paper. The moving obstacle avoidance algorithm is based on a Kalman filter through moving obstacle estimation and a Bezier curve for path generation. And, the network-based mobile robot, that is a unified system composed of distributed environmental sensors, mobile actuators, and controller, is compensated by a network delay compensation algorithm for degradation performance by network delay. The network delay compensation method by a sensor fusion using the Kalman filter is proposed for the localization of the robot to compensate both the delay of readings of an odometry and the delay of reading of environmental sensors. Through some simulation tests, the performance enhancement of the proposed algorithm in the viewpoint of efficient path generation and accurate goal point is shown here.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.49 no.10
• /
• pp.502-508
• /
• 2000

This paper presents a dynamic simulation algorithm for studying the effect of United Power Flow Controller(UPFC) on the low frequency power system oscillations and transient stability studies. The proposed algorithm is a Newton-type one and uses current injection type UPFC model, which gives a fast convergence characteristics. The algorithm is applied to studying inter-area power oscillation damping enhancement of a sample two-area power system both in time domain and frequency domain. The case study results show that the proposed algorithm is very efficient and UPFC is very effective and robust against operating point change.

• Communications for Statistical Applications and Methods
• /
• v.29 no.4
• /
• pp.441-451
• /
• 2022

In the era of bigdata, scalability is a crucial issue in learning models. Among many others, the Alternating Direction of Multipliers (ADMM, Boyd et al., 2011) algorithm has gained great popularity in solving large-scale problems efficiently. In this article, we propose applying the ADMM algorithm to solve the least square problem penalized by the pairwise-difference penalty, frequently used to identify group structures among coefficients. ADMM algorithm enables us to solve the high-dimensional problem efficiently in a unified fashion and thus allows us to employ several different types of penalty functions such as LASSO, Elastic Net, SCAD, and MCP for the penalized problem. Additionally, the ADMM algorithm naturally extends the algorithm to distributed computation and real-time updates, both desirable when dealing with large amounts of data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.10
• /
• pp.4903-4929
• /
• 2018

In this study, a fully automatic pose and expression invariant 3D face alignment algorithm is proposed to handle frontal and profile face images which is based on a two pass course to fine alignment strategy. The first pass of the algorithm coarsely aligns the face images to an intrinsic coordinate system (ICS) through a single 3D rotation and the second pass aligns them at fine level using a minimum nose tip-scanner distance (MNSD) approach. For facial recognition, multi-view faces are synthesized to exploit real 3D information and test the efficacy of the proposed system. Due to optimal separating hyper plane (OSH), Support Vector Machine (SVM) is employed in multi-view face verification (FV) task. In addition, a multi stage unified classifier based face identification (FI) algorithm is employed which combines results from seven base classifiers, two parallel face recognition algorithms and an exponential rank combiner, all in a hierarchical manner. The performance figures of the proposed methodology are corroborated by extensive experiments performed on four benchmark datasets: GavabDB, Bosphorus, UMB-DB and FRGC v2.0. Results show mark improvement in alignment accuracy and recognition rates. Moreover, a computational complexity analysis has been carried out for the proposed algorithm which reveals its superiority in terms of computational efficiency as well.