• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.17 no.2
• /
• pp.28-33
• /
• 2003

The piezoelectric transformer converters electrical energy into mechanical energy, It is high efficiency and small size transformer for high output voltage. The piezoelectric transformer operates the resonance frequency and the output voltage waveform is close to sine wave. Therefore, it is suitable for driving the LCD backlight in the notebook computer. In this paper, we discussed about the inverter which os driving piezoelectric transformer by generating sine wave through LC resonance after converting input DC voltage to the gate signal of FET. As the result of experiments, it was showed that the resonance frequency and voltage gain of the piezoelectric transformer was proportional to the load variation, and voltage gain was independent of the input voltage variation.

• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.484-490
• /
• 1999

품사 태그세트 매핑은 서로 다른 품사 태그세트로 태깅되어 있는 대량의 코퍼스들로부터 정보를 얻고 또한 제공함을 통해 코퍼스의 재사용성(reusability)을 높이는데 유용하게 사용된다. 본 논문은 포항공대 자연언어처리 연구실의 자연언어처리 엔진(SKOPE)의 품사 태거(POSTAG)에서 사용되는 태그세트와 한국전자통신연구원의 표준 태그세트 간의 양방향 태그세트 매핑을 다룬다. 매핑을 통해 표준태그세트로 태깅된 코퍼스로부터 POSTAG를 위한 대용량 학습자료를 얻고 POSTAG 가 두 가지 태그세트로 결과를 출력할 수 있다. 특히 한국어 태그세트 매핑에서 발생할 수 있는 여러 가지 문제점들, 즉 사전 표제어 차이 (형태소 분할 차이), 태그 할당 차이, 축약 처리 차이 등과 그것들의 기계적인 해결책을 살펴보고, 태그세트 매핑의 정확도를 측정하기 위해서 매핑 전과 후의 태깅 시스템의 정확도를 서로 비교함으로써 매핑의 정확도를 측정하는 실험을 수행하였다. 본 자동 매핑 방법을 반영한 POSTAG 는 제 1회 형태소 분석기 평가 대회(MATEC'99)에 적용되어 성공적으로 사용되었다.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2017.04a
• /
• pp.19-19
• /
• 2017

자율주행 트랙터를 위한 트랙터 조향제어는 일반적으로 전자모터를 이용한 EPS(Electric Power Steering) 시스템을 스티어링 휠에 연결하여 회전변위를 변경하고 그 결과 오비트롤(Orbitrol) 밸브의 토출유량을 바꾸고 호스로 연결된 조향실린더의 변위를 조절하여 최종적으로 전방 타이어의 방향각을 변경하면서 이루어진다. 이러한 조향방식은 시스템 구조상 조향실린더와 오비트롤 밸브가 상대적으로 멀리 떨어져 있으며, 밸브 특성상 약 ${\pm}5^{\circ}$의 오버랩이 포함되어 있다. 또한, EPS의 전자모터는 관성력, 마찰, 백래시 등의 영향을 가진다. 이와 같은 복합적인 영향은 조향 응답을 느리게 만들어 상대적으로 빠른 속도에서 주행에서 추종성능이 떨어지는 문제가 발생한다. 본 연구에서는 자율주행 트랙터의 조향성능 개선 연구의 일환으로 조향 HILS 시뮬레이터를 설계제작하여 조향 성능의 요인을 실험적으로 구명하고자 하였으며 이를 바탕으로 조향 시스템의 설계개선 방안을 수립하고자 하였다. 시뮬레이터는 동양물산 80 마력급 TX803 트랙터에 사용되는 오픈센터방식의 오비트롤 유압회로 시스템을 기어펌프가 장착된 AC모터로 구동되게 구성하였으며, 유량은 모터의 주파수를 조절 회전속도를 조절 변경하였다. 추가적으로 EPS와 오비트롤 조합의 조향성능을 비교 및 개선하기 위해 비례제어밸브(PVG 32, Danfoss)를 추가 장착하였다. 실제 트랙터 조향 시 나타나는 마찰저항을 모사하기 위해 부하 실린더를 구성하였으며, 조향 실린더의 부하의 크기는 부하 실린더를 폐회로를 구성하고 유량비례제어밸브를 이용한 유로의 개구량 조절을 통해 부하의 크기를 약 4000 N 까지 증가시킬 수 있도록 하였다. EPS와 비례제어밸브를 제어하기 위해 CANoe 8.0 소프트웨어를 이용하여 CAN통신 기반 가상 조향ECU를 구성하였으며 오비트롤의 기본 성능을 확인하기 위해 조향휠에 따른 실린더 동특성 및 계단 추종성능을 비례제어밸브와 비교하였다. 오비트롤 밸브는 약 ${\pm}5^{\circ}$이상 동작 시 실린더 압력이 상승하기 시작하였으며, 이후 약 ${\pm}10^{\circ}$이상 동작 시 조향실린더가 동작하기 시작하였다. 계단 추종성능실험에서는 비례제어밸브가 약 2배 이상의 응답개선을 나타냈다. 자율주행 경로추종 성능을 향상시키기 위해서는 순간적인 출력밀도가 높은 비례제어밸브를 통해 응답개선이 필요한 것으로 나타났다.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.40 no.3
• /
• pp.165-172
• /
• 2016

An experimental study of the thermal-hydraulic characteristics of passive safety systems (PSSs) was conducted using a system-integrated modular advanced reactor-integral test loop (SMART-ITL). The present passive safety injection system for the SMART-ITL consists of one train with the core makeup tank (CMT), the safety injection tank, and the automatic depressurization system. The objective of this study is to investigate the injection effect of the PSS on the small-break loss-of-coolant accident (SBLOCA) scenario for a 0.4 inch line break in the safety-injection system (SIS). The steady-state condition was maintained for 746 seconds before the break. When the major parameters of the target value and test results were compared, most of the thermal-hydraulic parameters agreed closely with each other. The water level of the reactor pressure vessel (RPV) was maintained higher than that of the fuel assembly plate during the transient, for the present CMT and safety injection tank (SIT) flow rate conditions. It can be seen that the capability of an emergency core cooling system is sufficient during the transient with SMART passive SISs.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.4
• /
• pp.487-494
• /
• 2023

In this study, performance comparison was performed on R and TensorFlow, which are free deep learning tools. In the experiment, six types of deep neural networks were built using each tool, and the neural networks were trained using the 10-year Korean temperature dataset. The number of nodes in the input layer of the constructed neural network was set to 10, the number of output layers was set to 5, and the hidden layer was set to 5, 10, and 20 to conduct experiments. The dataset includes 3600 temperature data collected from Gangnam-gu, Seoul from March 1, 2013 to March 29, 2023. For performance comparison, the future temperature was predicted for 5 days using the trained neural network, and the root mean square error (RMSE) value was measured using the predicted value and the actual value. Experiment results shows that when there was one hidden layer, the learning error of R was 0.04731176, and TensorFlow was measured at 0.06677193, and when there were two hidden layers, R was measured at 0.04782134 and TensorFlow was measured at 0.05799060. Overall, R was measured to have better performance. We tried to solve the difficulties in tool selection by providing quantitative performance information on the two tools to users who are new to machine learning.

• Radiation Oncology Journal
• /
• v.11 no.1
• /
• pp.175-181
• /
• 1993

Radiosurgery requires integral procedure where special devices and computer systems are needed for localization, dose planning and treatment. The aim of this work is to verify the overall mechanical accuracy of our LINAC and develop dose calculation algorithm for LINAC radiosurgery. The alignment of treatment machine and the performance testing of the entire system were extensively carried out and the basic data such as percent depth dose, off-axis ratio and output factor were measured. A three dimensional treatment planning system for stereotactic radiosurgery has been developed. We used an IBM personal computer with C programming language (IBM personal system/2, Model 80386, IBM Co., USA) for calculating the dose distribution. As a result, deviations at isocenter on gantry and table rotation for our treatment machine were acceptable since they were less than 2 mm. According to the phantom experiments, the focusing isocenter were successful by the error of less than 2 mm. Finally, the mechanical accuracy of our three dimensional planning system was confirmed by film dosimetry in sphere phantom.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.40 no.1
• /
• pp.9-16
• /
• 2016

A position control system composed of the PMLSM(Permanent Magnet Linear Synchronous Motor), unlike conventional linear permanent magnet synchronous motor is fixed to the permanent magnet moving coil rails (permanent magnet = stator, coil = mover), the coil is fixed, moving the permanent magnet, we propose a position control system (permanent magnet = mover, coil = stator) structure. Position is measured not using conventional encoder or resolver but by adopting vector control method using 2 hall sensors generating rectangular signal. This method estimate the velocity and position of mover by using the quadruple of two hall sensor signal instead of encoder signal. Vector control of PMLSM using 2 hall sensor generating rectangular wave is proved to control the system stable and efficiently through simulation. Also hardware experiment reveals that the position control performance is measured within the range of $30{\sim}50{\mu}m$ in the accuracy of $10{\sim}20{\mu}m$, which is improved twice to the conventional method. The proposed method exhibits its economical efficiency and practical usefulness. The vector control technique using two hall sensors can be installed in narrow place, accordingly it can be implemented on the system where the conventional encoder or resolver cannot operate.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.39 no.4
• /
• pp.351-358
• /
• 2015

In the recent times, the use of gasoline direct injection (GDI) engines has been regarded as a means of enhancing conformance to emission regulations and improving fuel efficiency. GDI engines have been widely adopted in the recent years for their better engine performance and fuel economy compared to those of conventional MPI gasoline engines. However, they present some disadvantages related to the mass and quantity of particulate matter generated during their use. This study investigated the nanoparticle characteristics of the particulate matter exhausted from a GDI engine vehicle installed with different types of gasoline particulate filters, after subjecting it to ultra-lean burn driving conditions. Three metal foam and metal fiber filters were used for each experimental condition. The number concentrations of particles were analyzed for understanding their behavior, and the reduction characteristics were obtained for each type of filter.

• The KIPS Transactions:PartB
• /
• v.19B no.3
• /
• pp.195-200
• /
• 2012

In Korean, spelling change in various forms must be recovered into base forms in morphological analysis as well as part-of-speech (POS) tagging is difficult without morphological analysis because Korean is agglutinative. This is one of notorious problems in Korean morphological analysis and has been solved by morpheme recovery rules, which generate morphological ambiguity resolved by POS tagging. In this paper, we propose a morpheme recovery scheme based on machine learning methods like Na$\ddot{i}$ve Bayes models. Input features of the models are the surrounding context of the syllable which the spelling change is occurred and categories of the models are the recovered syllables. The POS tagging system with the proposed model has demonstrated the $F_1$-score of 97.5% for the ETRI tree-tagged corpus. Thus it can be decided that the proposed model is very useful to handle morpheme recovery in Korean.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.38 no.9
• /
• pp.967-972
• /
• 2014

The good light permeability and hardness of glass allow it to be used in various fields. Non-conventional machining methods have been used for glass machining because of its brittle properties. As one non-contact machining method, a laser has advantages that include a high machining speed and the fact that no tool making is required. However, glass has light permeability. Thus, the use of a laser to machine glass has limitations. A nanosecond pulse laser can be used at low power for laser-induced backside wet etching, which is an indirect method. In previous studies, a short-wave laser that had good light absorption but a high price was used. In this study, a near-infrared laser was used to test the possibility of glass micro-machining. In particular, when deeper machining was conducted on a glass structure, more problems could result. To solve these problems, microstructure manufacturing was conducted using ultrasonic vibration.