• Journal of Electrical Engineering and Technology
• /
• v.10 no.1
• /
• pp.165-175
• /
• 2015

Energy storage system has been widely applied in power distribution sectors as well as in renewable energy sources to ensure uninterruptible power supply. This paper presents a model predictive algorithm to control a bidirectional AC-DC converter, which is used in an energy storage system for power transferring between the three-phase AC voltage supply and energy storage devices. This model predictive control (MPC) algorithm utilizes the discrete behavior of the converter and predicts the future variables of the system by defining cost functions for all possible switching states. Subsequently, the switching state that corresponds to the minimum cost function is selected for the next sampling period for firing the switches of the AC-DC converter. The proposed model predictive control scheme of the AC-DC converter allows bidirectional power flow with instantaneous mode change capability and fast dynamic response. The performance of the MPC controlled bidirectional AC-DC converter is simulated with MATLAB/Simulink(R) and further verified with 3.0kW experimental prototypes. Both the simulation and experimental results show that, the AC-DC converter is operated with unity power factor, acceptable THD (3.3% during rectifier mode and 3.5% during inverter mode) level of AC current and very low DC voltage ripple. Moreover, an efficiency comparison is performed between the proposed MPC and conventional VOC-based PWM controller of the bidirectional AC-DC converter which ensures the effectiveness of MPC controller.

• Journal of KIISE
• /
• v.44 no.3
• /
• pp.306-313
• /
• 2017

Named entity recognition (NER) seeks to locate and classify named entities in text into pre-defined categories such as names of persons, organizations, locations, expressions of times, etc. Recently, many state-of-the-art NER systems have been implemented with bidirectional LSTM CRFs. Deep learning models based on long short-term memory (LSTM) generally depend on word representations as input. In this paper, we propose an approach to expand word representation by using pre-trained word embedding, part of speech (POS) tag embedding, syllable embedding and named entity dictionary feature vectors. Our experiments show that the proposed approach creates useful word representations as an input of bidirectional LSTM CRFs. Our final presentation shows its efficacy to be 8.05%p higher than baseline NERs with only the pre-trained word embedding vector.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.7 no.1
• /
• pp.13-25
• /
• 2005

In general, tunnel ventilation is analyzed with the assumption of equal vehicular speed for both directions in bidirectional traffic tunnels. This practice is likely to result in minimizing the piston effects and cannot take into consideration the effects of real situations, since in most cases, speeds of vehicles moving in opposite directions are not equal. Therefore, The ultimate goal of this study is to review the effects of unequal vehicular speeds on the planning of local bidirectional tunnel ventilation. To apply unequal vehicular speed for the bidirectional tunnel ventilation plan, the following requirements are necessary; (1) Adoption of strict smoke concentration standards for 'free-flow & congested' traffic conditions, (2) Selection of an appropriate ratio of heavy directional traffic volume, and (3) Selection of reasonable stepwised magnitude of speed difference. Based on the importance of research topic, this study aims at comparing the differences of the capacity of ventilation equipment for the cases with equal and unequal vehicular speeds in local bidirectional tunnels.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 2006.07a
• /
• pp.743-743
• /
• 2006

• Korean Journal of Remote Sensing
• /
• v.38 no.3
• /
• pp.299-309
• /
• 2022

Recently, The sea water temperature around Korean Peninsula is steadily increasing. Water temperature changes not only affect the fishing ecosystem, but also are closely related to military operations in the sea. The purpose of this study is to suggest which model is more suitable for the field of water temperature prediction by attempting short-term water temperature prediction through various prediction models based on deep learning technology. The data used for prediction are water temperature data from the East Sea (Goseong, Yangyang, Gangneung, and Yeongdeok) from 2016 to 2020, which were observed through marine observation by the National Fisheries Research Institute. In addition, we use Long Short-Term Memory (LSTM), Bidirectional LSTM, and Gated Recurrent Unit (GRU) techniques that show excellent performance in predicting time series data as models for prediction. While the previous study used only LSTM, in this study, the prediction accuracy of each technique and the performance time were compared by applying various techniques in addition to LSTM. As a result of the study, it was confirmed that Bidirectional LSTM and GRU techniques had the least error between actual and predicted values at all observation points based on 1 hour prediction, and GRU was the fastest in learning time. Through this, it was confirmed that a method using Bidirectional LSTM was required for water temperature prediction to improve accuracy while reducing prediction errors. In areas that require real-time prediction in addition to accuracy, such as anti-submarine operations, it is judged that the method of using the GRU technique will be more appropriate.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.5
• /
• pp.410-415
• /
• 2013

This paper proposes a bidirectional platoon control law using a coupled distance error based on the backstepping-like feedback linearization control method for an interconnected mobile agent system with a string structure. Unlike the previous results where the single agent was controlled using the only own information without other agents, the proposed control law cannot show the only distance error convergence of each agent, but also the string stability of the whole system. Also, the control performances are improved by the proposed control law in spite of low performance of bidirectional control strategy in the previous results. The proposed bidirectional platoon control algorithm is based on the backstepping-like feedback linearization control method. The position errors between each agent and the preceding and the behind agents are coupled by weighted summation. By the proposed control law, the distance error of each agent can converge to zero while the string stability is guaranteed when the coupled errors can converge to zero. To this end, the back-stepping control method is employed. The pseudo velocity input is determined considering the kinematic relationship between agents and the string stability. Then, the actual dynamic control input is determined to make the actual velocity converge to the pseudo velocity input. The stability analysis and the simulation results of the proposed method are included in order to demonstrate the practical application of the proposed algorithm.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.24 no.1
• /
• pp.19-27
• /
• 2020

Reinforced concrete (RC) buildings built in the 1980s are vulnerable to seismic behavior because they were designed without any consideration of seismic loads. These buildings have widely spaced transverse reinforcements and a short lap splice length of longitudinal reinforcements, which makes them vulnerable to severe damage or even collapse during earthquakes. The purpose of this study is to investigate the impact of bidirectional lateral loads on RC columns with deficient reinforcement details. An experimental test was conducted for two full-scale RC column specimens. The test results of deficient RC columns revealed that bidirectional loading deteriorates the seismic capacity when compared with a column tested unidirectionally. Modeling parameters were extracted from the tested load-displacement response and compared with those proposed in performance-based design standards. The modeling parameters proposed in the standards underestimated the deformation capacity of tested specimens by nearly 50% and overestimated the strength capacity by 15 to 20%.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.24 no.6
• /
• pp.438-444
• /
• 2019

In this paper, bidirectional LLC resonant DC/DC converters with the primary auxiliary windings in transformers of resonant circuits are proposed. Although the resonant capacitors are used on both the primary and secondary sides, regardless of the direction of power flow, the main feature of the proposed converters exhibits high gain characteristics without any mutual coupling between the resonant capacitors. For one of the proposed converters, an investigation of the operating characteristics in each mode has been carried out. A prototype of a 3.3 kW bidirectional LLC resonant converter for interfacing 750 V DC buses has been built and tested to verify the validity and applicability of the proposed converter.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.4
• /
• pp.41-49
• /
• 2017

Word sense disambiguation(WSD) that determines the exact meaning of homonym which can be used in different meanings even in one form is very important to understand the semantical meaning of text document. Many recent researches on WSD have widely used NNLM(Neural Network Language Model) in which neural network is used to represent a document into vectors and to analyze its semantics. Among the previous WSD researches using NNLM, RNN(Recurrent Neural Network) model has better performance than other models because RNN model can reflect the occurrence order of words in addition to the word appearance information in a document. However, since RNN model uses only the forward order of word occurrences in a document, it is not able to reflect natural language's characteristics that later words can affect the meanings of the preceding words. In this paper, we propose a WSD scheme using Bidirectional RNN that can reflect not only the forward order but also the backward order of word occurrences in a document. From the experiments, the accuracy of the proposed model is higher than that of previous method using RNN. Hence, it is confirmed that bidirectional order information of word occurrences is useful for WSD in Korean language.

• Proceedings of the KIPE Conference
• /
• 2016.07a
• /
• pp.323-324
• /
• 2016

In this paper a high efficiency bidirectional resonant converterfor Vehicle-to-Grid applications (V2G) is proposed.The proposed converter has adopted an LC auxiliary circuit in the third winding of the transformer. With the proposed method full softswitching can be ensured in all switches over a wide range of loadsand the secondary ringing can be removed with no additional snubber or clamp circuitry.In addition, since the proposed resonant converter is able to operate at an almost constant resonant frequencyregardless of the load, CC/CV charge of the battery can be simply implemented with high efficiency. A 3.3 kW bidirectional converter for On-Board Charger of Electric Vehicle is implemented to verify the validity of the proposed method. The experimental results show the high efficiency characteristics of the proposed converter over the wide range of load in both charge and discharge mode. The maximum efficiency of the proposed system was 98.13 % at 2.3 kW during the constant voltage mode charge operation.