http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
자율주행 차량 운행을 위한 이상 의미 탐지 및 모니터링
윤한솔(Hansol Yoon) 국방로봇학회 2024 국방로봇학회 논문집 Vol.3 No.1
We investigate the problem of detecting anomalous behaviors of deployed autonomous vehicles that are not consistent with “typical” or “baseline” behaviors. Detecting anomalies requires us to process large volumes of high dimensional data from on-board sensors including GPS, camera, LiDAR and RADAR. In this work, we propose a simple methodology of runtime monitoring through “property templates” that allow us to detect meaningful anomalies of interest to stakeholders, while at the same time reducing from a very high dimensional space to a lower dimensional space through the property templates. Next, we use Pareto frontiers to define anomalous instances in the data with respect to the property. We conduct a preliminary evaluation of our property template approach using data collected from a pilot autonomous shuttle vehicle deployment. We demonstrate our approach through a set of standard property templates that seek to specify common safety-critical behaviors around maintaining following distance to a lead vehicle, navigating through stop signs, and operating under a speed limit. Using this case-study, we examine the anomalies detected in detail, and discuss some practical challenges for real-time anomaly monitoring.
조철민(Chulmin Cho),윤한솔(Hansol Yoon),윤헌준(Heonjun Yoon),김홍진(Hongjin Kim),이소원(Sowon Lee),윤병동(Byeng Dong Youn),김재은(Jae Eun Kim),김윤영(Yoon Young Kim) 대한기계학회 2012 대한기계학회 춘추학술대회 Vol.2012 No.11
Prior to designing energy harvesters, it is important to understand ambient energy characteristics and preliminary computation of harvestable energy. With special attention to vibrational and thermal ambient energy sources, this study aims at developing a corresponding energy map that can visualize ambient and harvestable energy). The energy map benefits energy harvesting technology in many ways. First, economic feasibility can be studied while selecting best sites for energy harvesting. Second, the map facilitates the conceptual and detailed design of an energy harvester type (e.g., piezoelectric, thermoelectric) and its target application such as wireless sensor network. The three-fold steps are proposed to build the energy map: 1) data acquisition for available energy, 2) harvestable energy analysis with energy conversion models, and 3) visualization of harvestable energy. We obtain site-specific harvestable energy maps and open discussion about the benefits of the proposed idea.
전기 방사 공정의 제조 변수를 이용한 TiO<sub>2</sub> 나노섬유의 직경 제어
윤한솔,김보성,김완태,나경한,이정우,양완희,박동철,최원열,Yoon, Han-Sol,Kim, Bo-Sung,Kim, Wan-Tae,Na, Kyeong-Han,Lee, Jung-Woo,Yang, Wan-Hee,Park, Dong-Cheol,Choi, Won-Youl 한국결정성장학회 2021 한국결정성장학회지 Vol.31 No.1
TiO2는 물리적, 화학적 안정성이 높고, 신체에 무해하여 태양전지, 치과용 임플란트 및 광촉매 같은 다양한 분야에서 사용되어 왔다. 비표면적이 큰 TiO2 나노섬유는 생체 친화성 제품에서 좋은 반응성과 공기 및 수질 정화시 우수한 광촉매 특성을 보여주었다. TiO2 나노섬유를 제조하기 위해 전기방사법을 사용하였으며, 제조 변수에 따른 직경 변화를 관찰하기 위해 precursor 성분 변수와 공정 변수로 구분하여 미세구조 변화를 분석하였다. Precursor 성분 변수로는 PVP(Polyvinylpyrrolidone) 및 TTIP(Titanium(IV) isopropoxide)의 농도를 선택하였고, 공정 변수로는 주입 속도와 인가 전압을 선택하였다. TiO2 나노섬유의 미세구조와 결정구조는 FE-SEM(Field emission scanning electron microscope)와 XRD(X-ray diffraction)을 이용하여 분석하였다. 450℃에서 3시간 열처리 공정을 통해, 평균 직경 약 0.27 ㎛에서 1.31 ㎛를 갖는 asspun TiO2 나노섬유가 0.22 ㎛에서부터 0.78 ㎛의 평균 직경을 갖는 anatase 상의 TiO2 나노섬유로 상전이 됨을 확인할 수 있었다. 평균 직경 0.22 ㎛의 anatase TiO2 나노섬유는 비 표면적 증대에 의한 광촉매 특성 향상을 기대할 수 있다. 또한 TiO2 나노섬유의 직경 변화를 위해서는 주입 속도 및 인가 전압과 같은 공정 변수보다는 PVP 농도 및 TTIP 농도와 같은 precursor 성분 변수를 제어하는 것이 더욱 효과적이었다. TiO2 has been used in various fields such as solar cells, dental implants, and photocatalysis, because it has high physical and chemical stability and is harmless to the body. TiO2 nanofibers which have a large specific surface area also show a good reactivity in bio-friendly products and excellent photocatalysis in air and water purification. To fabricate TiO2 nanofibers, an electrospinning method was used. To observe the diameter of TiO2 nanofibers with fabrication variables, the fabrication variables was divided into precursor composition variables and process variables and microstructure was analyzed. The concentrations of PVP (Polyvinylpyrrolidone) and TTIP (Titanium(IV) isopropoxide) were selected as precursor composition variables, and inflow velocity and voltage were also selected as process variables. Microstructure and crystal structure of TiO2 nanofibers were analyzed using FE-SEM (Field emission scanning electron microscope) and XRD (X-ray diffraction), respectively. As-spun TiO2 nanofibers with an average diameter of about 0.27 ㎛ to 1.31 ㎛ were transformed to anatase TiO2 nanofibers with an average diameter of about 0.22 ㎛ to 0.78 ㎛ after heat treatment of 3 hours at 450℃. Anatase TiO2 nanofibers with an average diameter of 0.22 ㎛ can be expected to improve the photocatalytic properties by increasing the specific surface area. To change the average diameter of TiO2 nanofibers, the control of precursor composition variables such as concentrations of PVP and TTIP is more efficient than the control of electrospinning process variables such as inflow velocity and voltage.
강의평가 주관식 응답에 나타난 교수와 학생의 의견 차이 비교 분석: 강의의 장‧단점 의견에 대한 네트워크 및 내용분석
민혜리(Hyeree, Min),윤한솔(Hansol, Yoon) 학습자중심교과교육학회 2017 학습자중심교과교육연구 Vol.17 No.11
본 연구는 강의에 대한 교수의 자기평가와 학생의 평가를 비교하기 위해, 교수와 학생의 동일 강의에 대한 주관식 응답결과를 분석하여 교수집단과 학생집단의 인식 차이, 단과대학별 특성 등을 살펴보고자 하였다. 2013년부터 2016년까지 동일 문항의 강의평가에 참여한 교수133명, 학생 4665명의 강의평가 주관식 응답자료를 6개 단과 대학으로 나누어 분석하였다. 분석방법은 네트워크분석방법으로 고려대 김규태 교수 연구팀이 제작한 엑셀 매크로 프로그램을 이용하여 키워드와 워드클라우드 이미 지를 추출하고, 키워드 네트워크를 Ucinet과 Netdraw로 분석하였다. 추출한 키워드는 다시 최종 8개 범주(강의환경, 강의자료, 강의운영, 설명방식, 상호작용, 교수특성, 평가, 기타)로 분류하였다. 연구결과, 교수와 학생 모두 설명방식에 대한 언급이 가장 많았는데 두 집단 모두수업을 평가할 때 가장 많이 고려하는 부분이 설명방식이 며, 학생의 수준과 이해도를 고려한 설명에 비중을 두어야 함을 시사한다. 또한 강의 자료와 교수특성, 강의운영 영역의 분석결과는 교수가 학생의 의견에 조금만 귀를 기울여도 강의개선이 가능함을 시사한다. 단과대별로 살펴보면 자연대에서는 교수-학생 그룹 간 의견의 일치도가 가장 낮게 나타났다. 강의개선을 위한 방향 설정에서양 집단의 인식차이를 파악하는 노력이 필요하다는 것을 의미한다 This study analyzed the result of response to the comment/essay survey on the same lecture between for both professors and students to compare self rating by professor on his or her lecture with student evaluation in order to find out the difference in recognition between professor group and student group and the characteristic of college etc. The lecture evaluation with same questions by 133 professors and the responses to the comment/essay survey on lecture by 4665 students that were conducted respectively from 2013 to 2016 were analyzed by dividing 6 colleges. For the method of analysis, Excel macro program produced by Prof. Kim Kyu-tae’s research team was used as network analysis method to extract key words and word cloud image and the keyword network was analyzed by Ucinet and Netdraw. The extracted keywords were classified as the final 8 categories: lecture environment, lecture material, lesson management, how to explain, interaction, the characteristic of professor, evaluation and others. The result showed that there were greatest number of comments in How to explain for both professor and student group and these both groups take account of How to explain most. In other words, it implies that it should put emphasis on explanation that considers the level of students and understanding. In addition, the analyzed result of lecture environment, the characteristic of professor and lesson management suggests that it’s possible to improve the quality of lecture if the professors give careful attention to the opinion of students. To review it by each college, the college of natural sciences showed lowest consistency for the opinions between professors and students. It means that it’s necessary to make efforts to understand the gap in recognition between both groups in terms of setting direction for lecture improvement.