• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.257-257
• /
• 2000

The objective of this research is to monitor and control the vehicle motion in order to remove out the existing safety risk based upon the human-machine cooperative vehicle control. A predictive control method is proposed to control the steering wheel of the vehicle to keep the lane. Desired angle of the steering wheel to control the vehicle motion could be calculated based upon vehicle dynamics, current and estimated pose of the vehicle every sample steps. The vehicle pose and the road curvature were calculated by geometrically fusing sensor data from camera image, tachometer and steering wheel encoder though the Perception Net, where not only the state variables, but also the corresponding uncertainties were propagated in forward and backward direction in such a way to satisfy the given constraint condition, maintain consistency, reduce the uncertainties, and guarantee robustness. A series of experiments was conducted to evaluate the control performance, in which a car Like robot was utilized to quit unwanted safety problem. As the results, the robot was keeping very well a given lane with arbitrary shape at moderate speed.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2018년도 추계학술대회
• /
• pp.81-83
• /
• 2018

Fully Convolutional Network(FCN)은 기존의 방법보다 뛰어난 성능을 보였지만, FCN은 RGB 정보만을 사용하기 때문에 세밀한 예측이 필요한 장면에서는 다소 부족한 성능을 보였다. 이를 해결하기 위해 인코더-디코더 구조를 이용하여 RGB와 깊이의 멀티 모달을 활용하기 위한 FuseNet이 제안되었다. 하지만, FuseNet에서는 RGB와 깊이 브랜치 사이의 융합은 있지만, 인코더와 디코더 사이의 특징 지도를 융합하지 않는다. 본 논문에서는 FCN의 디코더 부분의 업샘플링 과정에서 이전 계층의 결과와 2배 업샘플링한 결과를 융합하는 스킵 레이어를 적용하여 FuseNet의 모달리티를 잘 활용하여 성능을 개선했다. 본 실험에서는 NYUDv2와 SUNRGBD 데이터 셋을 사용했으며, 전체 정확도는 각각 77%, 65%이고, 평균 IoU는 47.4%, 26.9%, 평균 정확도는 67.7%, 41%의 성능을 보였다.

• 방송공학회논문지
• /
• 제27권5호
• /
• pp.794-807
• /
• 2022

동물보호법 개정에 따라 반려견 등록이 의무화 되었음에도 불구하고, 현재 등록 방법의 불편함으로 등록율이 저조한 상태이다. 본 논문에서는 새로운 등록 방법으로 검토되고 있는 반려견 안면 인식 기술에 대한 성능 개선 연구를 진행하였다. 딥러닝 학습을 통해, 반려견의 안면 인식을 위한 임베딩 벡터를 생성하여 반려견 개체별로 식별하기 위한 방법을 실험하였다. 딥러닝 학습을 위한 반려견 이미지 데이터셋을 구축하고, InceptionNet과 ResNet-50을 백본 네트워크로 사용하여 실험하였다. 삼중항 손실 방법으로 학습하였으며, 안면 검증과 안면 식별로 나뉘어 실험하였다. ResNet-50 기반의 모델에서 최고 93.46%의 안면 검증 성능을 얻을 수 있었으며, 안면 식별 시험에서는 rank-5에서 91.44%의 최고 성능을 각각 얻을 수 있었다. 본 논문에서 제시한 실험 방법과 결과는 반려견의 등록 여부 확인, 반려견 출입시설에서의 개체 확인 등 다양한 분야로 활용이 가능하다.

• Nuclear Engineering and Technology
• /
• 제40권6호
• /
• pp.467-476
• /
• 2008

To keep the superconducting (SC) magnet coils of KSTAR at proper operating conditions, not only the coils but also other cold components, such as thermal shields (TS), magnet structures, SC bus-lines (BL), and current leads (CL) must be maintained at their respective cryogenic temperatures. A helium refrigeration system (RRS) with an exergetic equivalent cooling power of 9 kW at 4.5 K without liquid nitrogen ($LN_2$) pre-cooling has been manufactured and installed. The main components of the KST AR helium refrigeration system (HRS) can be classified into the warm compression system (WCS) and the cryogenic devices according to the operating temperature levels. The process helium is compressed from 1 bar to 22 bar passing through the WCS and is supplied to cryogenic devices. The main components of cryogenic devices are consist of cold box (C/B) and distribution box (D/B). The C/B cool-down and make the various cryogenic helium for the KSTAR Tokamak and the various cryogenic helium is distributed by the D/B as per the KSTAR requirement. In this proceeding, we will present the commissioning results of the KSTAR HRS. Circuits which can simulate the thermal loads and pressure drops corresponding to the cooling channels of each cold component of KSTAR have been integrated into the helium distribution system of the HRS. Using those circuits, the performance and the capability of the HRS, to fulfill the mission of establishing the appropriate operating condition for the KSTAR SC magnet coils, have been successfully demonstrated.

• Nuclear Engineering and Technology
• /
• 제41권5호
• /
• pp.709-714
• /
• 2009

An ion implanter, which can serve as a metal-ion supply, has been constructed and performance tested. Copper ions are generated and extracted from a Bernas ion source with a heating crucible that provides feed gases to sustain the plasma. Sable arc plasmas can be sustained in the ion source for a crucible temperature in excess of $350^{\circ}C$. Stable extraction of the ions is possible for arc Currents less than 0.3 A. Arc currents increase with the induced power of a block cathode and the transverse field in the ion source. $Cu^+$ ions in the extracted beam are separated using a dipole magnet. A $20{\mu}A$ $Cu^+$ ion current can be extracted with a 0.2 A arc current. The ion current can support a dose of $10^{16}ions/cm^2$ over an area of $15\;cm^2$ within a few hours.

• Nuclear Engineering and Technology
• /
• 제51권7호
• /
• pp.1860-1865
• /
• 2019

The confinement degradation of the energetic particles during RMP would be a key issue in success of realizing the successful energy production using fusion plasma, because a 3.5 MeV energetic alpha particle should be able to sustain the burning plasma after the ignition. As KSTAR recent results indicate the generation of high-performance plasma(${\beta}_p{\sim}3$), the confinement of the energetic particles is also an important key aspect in neutral beam driven plasma. In general, the measured absolute value of the neutron intensity is generally used for to estimating the confinement time of energetic particles by comparing it with the theoretical value based on transport calculations. However, the availability of, but for its calculation process, many accurate diagnostic data of plasma parameters such as thermal and incident fast ion density, are essential to the calculation process. In this paper, the time evolution of the neutron signal from an He3 counter during the beam blank has permitted to facilitate the estimation of the slowing down time of energetic particles and the method is applied to investigate the fast ion effect on ELM suppressed KSTAR plasma which is heated by high energy deuterium neutral beams.

• Nuclear Engineering and Technology
• /
• 제53권6호
• /
• pp.1736-1746
• /
• 2021

The results of full-scale numerical experiments of a hybrid thorium-containing fuel cell facility operating in a close-to-critical state due to a controlled source of fusion neutrons are discussed in this work. The facility under study was a complex consisting of two blocks. The first block was based on the concept of a high-temperature gas-cooled thorium reactor core. The second block was an axially symmetrical extended plasma generator of additional neutrons that was placed in the near-axial zone of the facility blanket. The calculated models of the blanket and the plasma generator of D-T neutrons created within the work allowed for research of the neutronic parameters of the facility in stationary and pulse-periodic operation modes. This research will make it possible to construct a safe facility and investigate the properties of thorium fuel, which can be continuously used in the epithermal spectrum of the considered hybrid fusion-fission reactor.

• Nuclear Engineering and Technology
• /
• 제54권10호
• /
• pp.3703-3716
• /
• 2022

Target alignment technology is one of the most critical technologies in laser fusion experiments and is an important technology related to the success of laser fusion experiments. In this study, by combining the open-loop and closed-loop errors of the target alignment, the Kalman state observer is used to estimate the position of the target, which improves the observation precision of the target alignment. Then the optimized result is used to guide the alignment of the target. This method can greatly optimize the target alignment error and reduce uncertainty. With the improvement of the target alignment precision, it will greatly improve the reliability and repeatability of the experiments' results, thereby improving the success rate of the experiments.

• Nuclear Engineering and Technology
• /
• 제40권6호
• /
• pp.451-458
• /
• 2008

After more than 10 years construction, KSTAR (Korea Superconducting Tokamak Advanced Research) had finally completed its assembly in June 2007, and then achieved the goal of first-plasma in July 2008 through the four month's commissioning. KSTAR was constructed with fully superconducting magnets with material of $Nb_3Sn$ and NbTi, and their operation temperatures are maintained below 4.5K by the help of Helium Refrigerator System. During the first-plasma operation, plasmas of maximum current of 133kA and maximum pulse width of 865ms were obtained. The KSTAR Integrated Control System (KICS) has successfully fulfilled its missions of surveillance, device operation, machine protection interlock, and data acquisition and management. These and more were all KSTAR commissioning requirements. For reliable and safe operation of KSTAR, 17 local control systems were developed. Those systems must be integrated into the logically single control system, and operate regardless of their platforms and location installed. In order to meet these requirements, KICS was developed as a network-based distributed system and adopted a new framework, named as EPICS (Experimental Physics and Industrial Control System). Also, KICS has some features in KSTAR operation. It performs not only 24 hour continuous plant operation, but the shot-based real-time feedback control by exchanging the initiatives of operation between a central controller and a plasma control system in accordance with the operation sequence. For the diagnosis and analysis of plasma, 11 types of diagnostic system were implemented in KSTAR, and the acquired data from them were archived using MDSpius (Model Driven System), which is widely used in data management of fusion control systems. This paper will cover the design and implementation of the KSTAR integrated control system and the data management and visualization systems. Commissioning results will be introduced in brief.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제12권12호
• /
• pp.525-534
• /
• 2023

가스 누출 감지 시스템은 가스의 폭발성과 독성으로 인한 인명 피해를 최소화할 핵심적인 장치이다. 누출 감지 시스템은 대부분 단일 센서를 활용한 방식으로, 가스 센서나 열화상 카메라를 통한 검출 방식으로 진행되고 있다. 이러한 단일 센서 활용의 가스 누출감지 시스템 성능을 고도화하기 위하여, 본 연구에서는 가스 센서와 열화상 이미지 데이터에 멀티모달형 딥러닝을 적용한 연구를 소개한다. 멀티모달 공인 데이터셋인 MultimodalGasData를 통해 기존 논문과의 성능을 비교하였고, 가스 센서와 열화상 카메라의 단일모달 모델을 기반하여 네 가지 멀티모달 모델을 설계 및 학습하였다. 이를 통해 가스 센서와 열화상 카메라는 각각 1D CNN, GasNet 모델이 96.3%와 96.4%의 가장 높은 성능을 보였다. 앞선 두 단일모달 모델을 기반한 Early Fusion 형식의 멀티모달 모델 성능은 99.3%로 가장 높았으며, 또한 기존 논문의 멀티모달 모델 대비 3.3% 높았다. 본 연구의 높은 신뢰성을 갖춘 가스 누출 감지 시스템을 통해 가스 누출로 인한 추가적인 피해가 최소화되길 기대한다.