http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
강호성 ( Hoseong Kang ),서영진 ( Young Jin Seo ),고지은 ( Ji Eun Ko ),박치훈 ( Chi Hoon Park ) 한국공업화학회 2022 공업화학전망 Vol.25 No.5
최근들어 머신러닝 기술이 다양한 분야에서 획기적인 성과를 내며 많은 주목을 받고 있다. 이러한 발전에는 컴퓨터 기술의 급격한 발달과 딥러닝으로 대표되는 다양한 알고리즘의 등장이 그 배경에 있다. 이 중 컴퓨터 기술의 발달은 또 다른 컴퓨터 활용 분야인 전산모사 분야에도 큰 영향을 미쳐 기존에는 계산하기 어려웠던 크고 복잡한 구조의 모사가 가능해 졌고, 이에 따라 전산모사 관련 연구들이 활발히 진행되고 있다. 두 분야 모두 컴퓨터에 기반하고 있기 때문에, 최근 들어 이들 두 분야를 접목하여 전산모사-머신러닝 융합연구를 진행하고자 하는 시도들이 나타나고 있다. 이중에는 데이터베이스에 기반한 머신러닝 기법을 바탕으로 소재 개발을 수행하던 연구분야에서 기존에 저장된 데이터베이스가 아닌 전산모사에서 얻어진 결과를 입력 변수로 활용하고자 하는 시도들이 있고, 반대로 머신러닝을 통해서 얻어진 결과들을 전산모사 계산에 입력 변수로 활용하기 위한 시도들도 있다. 본 총설에서는 우선 에너지 소재 분야에 초점을 맞춰서 관련 전산모사 기술을 소개하고, 현재 진행되고 있는 전산모사-머신러닝 융합 연구 동향에 대하여 소개하고자 한다.
Kang, Minji,Seo, Jang Kyun,Choi, Hoseong,Choi, Hong Soo,Kim, Kook Hyung The Korean Society of Plant Pathology 2016 Plant Pathology Journal Vol.32 No.1
The infectious full-length cDNA clone of zucchini yellow mosaic virus (ZYMV) isolate PA (pZYMV-PA), which was isolated from pumpkin, was constructed by utilizing viral transcription and processing signals to produce infectious in vivo transcripts. Simple rub-inoculation of plasmid DNAs of pZYMV-PA was successful to cause infection of zucchini plants (Cucurbita pepo L.). We further engineered this infectious cDNA clone of ZYMV as a viral vector for systemic expression of heterologous proteins in cucurbits. We successfully expressed two reporter genes including gfp and bar in zucchini plants by simple rub-inoculation of plasmid DNAs of the ZYMV-based expression constructs. Our method of the ZYMV-based viral vector in association with the simple rub-inoculation provides an easy and rapid approach for introduction and evaluation of heterologous genes in cucurbits.
Cooperative Aerial Manipulation Using Multirotors With Multi-DOF Robotic Arms
Kim, Suseong,Seo, Hoseong,Shin, Jongho,Kim, H. Jin IEEE 2018 IEEE/ASME transactions on mechatronics Vol.23 No.2
<P>This paper investigates pose manipulation of a rod-shaped object using two aerial manipulators, which are multirotors combined with multi-degrees of freedom robotic arms. To achieve the cooperative aerial manipulation, we present motion control and planning methods for aerial manipulators. First, a robust multirotor motion controller is designed based on the thorough analysis on the dynamics of the aerial manipulator. By utilizing extended high-gain observer and disturbance observer techniques, each multirotor can be controlled regardless of the disturbances generated by the robotic arm, the object, and the peer aerial manipulator. Second, a motion planner that assures safety during the cooperative aerial manipulation is proposed. To achieve safe manipulation, we estimate and regulate the internal force between the aerial manipulators. In addition, an unilateral constraint is designed which imposes collision avoidance between the object and aerial manipulators. All the designed features are prioritized and transformed into the reference velocity of the aerial manipulators. The proposed control and guidance laws are validated by a successful autonomous aerial manipulation experiment.</P>
Model Predictive Control for an Aerial Manipulator Opening a Hinged Door
Dongjae Lee,Dohyun Jang,Hoseong Seo,H.Jin Kim 제어로봇시스템학회 2019 제어로봇시스템학회 국제학술대회 논문집 Vol.2019 No.10
Aerial manipulation has been widely studied to be employed in various tasks such as exploration and transportation. To incorporate aerial manipulation into more sophisticated tasks like pulling or pushing a heavy cargo, an active interaction with surrounding structures should be considered. Unlike physical contact with a static structure which was mainly studied in previous papers, interaction with a movable structure requires a consideration of dynamics of the structure which makes the scenario more complex. In this paper, an aerial manipulator opening a hinged door is presented. Coupled dynamics between an aerial manipulator and a hinged door is derived, and a model predictive control (MPC) algorithm using iterative Linear Quadratic Regulator (iLQR) method for the derived dynamic equation is proposed. Through our proposed control strategy, sub-optimal state and input trajectories robust to model uncertainties while satisfying input constraints are generated. Our dynamic model and control algorithm are validated through simulations.