http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이산 시간 슬라이딩 모드 제어 적용시 산업용 서보 시스템에서 사전 실험 및 오프라인 튜닝이 없는 보조 상태 변수 게인의 실시간 온라인 설정 방법
양대영,한지석,오태호,김영석,김태훈,이상훈,조동일 제어로봇시스템학회 2021 제어로봇시스템학회 국내학술대회 논문집 Vol.2021 No.6
This paper develops a real-time, on-line gain tuning methodology for discrete-time Sliding mode control with Decoupled disturbance compensator and Auxiliary state (SDA) under control input saturation without pre-tests and offline tuning. The proposed methodology is implemented in industrial servo systems and the experiment results are presented in this paper. The discrete-time SDA method is a robust control method dependent on control input saturation and disturbances. However, when the gain parameter of the auxiliary state is improperly tuned, it is hard to achieve desired performance specifications due to oscillations or repeated occurrences of control input saturation. Previously, a gain tuning algorithm for the auxiliary state using pre-tests and offline tuning has been proposed. This paper extends this previous gain tuning algorithm and proposes a new algorithm which does not require pre-tests and offline tuning by predicting the maximum position error based on a given reference profile and the control hardware input limit. The developed algorithm is implemented and experimental results show the improvements in the tuning process without the time-consuming offline tuning procedure.
양대영,이준호,최경열,최우근,박순호 사단법인 한국융합기술연구학회 2023 아시아태평양융합연구교류논문지 Vol.9 No.1
With continuous growth in the development and adoption of autonomous technologies in maritime trade worldwide, accurate assessment of ports and Container Terminals (CTs) play a significant role in improving the maritime business. However, the existing paradigm in CT performance analysis is limited to using simulations or focusing on a specific set of metrics and CTs. We propose a performance analysis method for CTs based on data collected with Electronic Data Interchange (EDI), the standard data exchange method in the maritime business. By processing the data generated from EDI and calculating metrics based on prior work (traffic volume, productivity index, vessel wait time, terminal wait time, truck turn time), the framework provides a broad view of CT performance in the ports of Korea. Due to the low data integrity of the raw data, the proposed method incorporates an Extract, Transform, Load (ETL) program to process the raw data into metric-specific data sets. The data sets then provide more accurate information on performances. We apply our method to real-world data collected from 2016 to 2021 to provide general insights and observe a 5% error when cross-referenced with external sources. Finally, our findings suggest the need for distinguishing analysis of performance according to the type of trade process (import, export, transshipment). 전 세계 항만들이 자동화 기술을 적극 도입하고 있는 만큼 국내 컨테이너 터미널의 동향 파악 및 성능 분석이 필요하지만 현존하는 패러다임은 단순화된 시뮬레이션을 사용하거나 특정 터미널 및 지표에 국한해 설계함으로써 제한적이라고 볼 수 있다. 이를 극복하기 위해서는 운송 중 생성되는 전자 데이터 교환(EDI) 데이터의 활용을 고려해 볼 수 있다. 본 연구는 EDI로부터 집계된 데이터를 정제하고 이를 통해 여러 성능 지표(물동량, 하역 생산성, 선박 대기 시간, 터미널 대기 시간, 트럭 턴 타임)를 계산하여 한국 항만 내 터미널을 분석하는 방안을 연구한다. 이를 위해 본 연구는 추출, 변환, 적재(ETL) 프로그램을 개발하였으며, 2016년부터 2021년까지 축적된 실제 데이터에 적용하였다. 이를 분석 시 선행연구 및 외부 자료와 비교하여 약 5% 오차율을 기록하였는데, 이는 본 연구의 유효성을 제시한다. 추가적으로, 해당 통계를 통해 운송 중 화물 종류(수출, 수입, 환적) 구분의 중요성에 대한 분석을 제공하였다.
배현미,양대영 木浦大學校 沿岸環境硏究所 1999 沿岸環境硏究 Vol.16 No.1
This study is to attempt in order to make the space as an identity and amenity and more intimacy of our cities by the preparation of Community Sign. This study aims to propose the direction of solve a problem, based on the analysis of the status on the sign of urban which is the need of sign facilities of urban is recently felt. The results clearly research of this study are as follows: ① Continuity allowance of furnish information and to avoid the disproportion of information. ② The necessity of research for a location and an establishment method of urban sign. ③ Quality support of Community Sign. ④ The introduction necessity of Community Sign can be communicate to everybody including the physically handicapped.
탄성파반사자료(彈性波反射資料)에서 지하물질(地下物質)의 음향계수(音響係數) 추출(抽出)
양승진,안대영,Yang, Sung-Jin,Ahn, Dae-Young 대한자원환경지질학회 1986 자원환경지질 Vol.19 No.4
Accoustic impedances of subsurface layers are determined from the amplitudes of reflections from the layers. Densities of the layers can also be calculated from the accoustic impedances in case where velocities are known by velocity analysis of reflection data or any other method. The accoustic impedance is a good information for direct identification of the kind of some subsurface material like coal or oil.
소규모 SAR 데이터셋에 효과적인 딥러닝 표적 식별 모델 탐구
양유경,최여름,채대영,송정민 한국정보기술학회 2025 한국정보기술학회논문지 Vol.23 No.3
SAR(Synthetic Aperture Radar) 영상은 기상에 의존하지 않고 원거리 획득이 가능한 장점으로 인해 군의 감시정찰 분야나 국가 재난 감시 등 다양한 분야에 활용되고 있다. 하지만 SAR 영상은 데이터 획득 비용 및 획득 소요 시간으로 인해 큰 규모의 데이터셋 구축이 어려워 딥러닝 기반 표적 식별 모델의 성능향상을 어렵게 한다. 이러한 문제를 해결하기 위해 시뮬레이션 SAR영상을 생성하여 활용하거나, 대규모 광학 영상 데이터셋에서 학습된 모델을 소규모 SAR 영상에 전이학습하여 사용한다. 본 논문에서는 대규모 광학 영상 기반으로 개발된 최신의 고성능 식별 모델들을 선정하고 시뮬레이션 SAR 영상 및 소규모 SAR 실측 데이터셋에 전이학습하여 성능을 비교 분석함으로써 소규모 SAR 데이터셋에 효과적인 표적 식별 모델을 탐구한다. SAR(Synthetic Aperture Radar) images are used in various fields such as military surveillance and reconnaissance and national disaster monitoring due to the advantage of being able to acquire images from a long distance without relying on the weather condition. However, it is difficult to build a large-scale datasets due to data acquisition costs and time, making it difficult to improve the performance of deep learning-based target recognition models. To solve this problem, simulated SAR images are often utilized, and there is a method of transferring a model learned from a large-scale optical images to a small-scale SAR images. In this paper, we select the latest state-of-the-art recognition models developed based on large-scale optical images and explore effective target recognition models for small-scale SAR datasets by comparing and analyzing their performance through transfer learning on simulated SAR images and small-scale SAR real images.
유종영,양대영,신희동,손일,Yu, Jong Yeong,Yang, Dae Young,Shin, Hee Dong,Sohn, Il 한국자원리싸이클링학회 2015 자원리싸이클링 Vol.24 No.3
Raw materials in steel industry decide on the productivity, quality and price competitiveness. Utilizing iron-containing by-products as raw materials for steel products can save the cost of cleaning up iron-containing by-products and solve environmental issues. Iron-containing by-products have a small particle size. If they are directly inserted in a steel making process, it cause a problem such as poor heat flow and scattering. To solve these problems and induce the additional reduction, study concern with iron ore-coal mixed briquette technique are conducted by many researchers. In this paper, method of making carbon composite briquette iron(CCBI) using iron-containing by-products was studied. The effect of composition of Fe-containing process wastes, reducing agent, flux and binder on formability of CCBI (carbon composite briquette iron) was measured.