• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.552-557
• /
• 2004

This paper presents a disturbance observer based on an $H_{\infty}$ controller synthesis for the trajectory control of a hydraulic excavator. Compared to conventional robot manipulators driven by electrical motors, the hydraulic excavator has more nonlinear and coupled dynamics. In particular, the interactions between an excavation tool and the materials being excavated are unstructured and complex. In addition, its operating modes depend on working conditions, which make it difficult to not only derive the exact mathematical model but also design a controller systematically. In this study, the approximated linear model obtained through off-line system identification is used as nominal plant model for a disturbance observer. A disturbance observer based tracking controller which considers the effect of disturbance and model uncertainty is synthesized in $H_{\infty}$ frameworks. Simulation results are used to demonstrate the applicability of the proposed control scheme.

• 한국정보시스템학회:학술대회논문집
• /
• 한국정보시스템학회 1996년도 추계학술발표회 발표논문집
• /
• pp.217-227
• /
• 1996

The new field of learning control develops controllers that learn to improve their performance at executing a given task, based on experience performin this specific task. In a previous work［6］, the authors presented a theory of indirect learning control based on use of indirect adaptive control concepts employing simultaneous identification ad control. This paper develops improved indirect learning control algorithms, and studies the use of such controllers in decentralized systems. The original motivation of the learning control field was learning in robots doing repetitive tasks such as on an assembly line. This paper starts with decentralized discrete time systems, and progresses to the robot application, modeling the robot as a time varying linear system in the neighborhood of the nominal trajectory, and using the usual robot controllers that are decentralized, treating each link as if it is independent of any coupling with other links. The basic result of the paper is to show that stability of the indirect learning controllers for all subsystems when the coupling between subsystems is turned off, assures convergence to zero tracking error of the decentralized indirect learning control of the coupled system, provided that the sample time in the digital learning controller is sufficiently short.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2004년도 춘계학술대회 논문집
• /
• pp.49-54
• /
• 2004

유비쿼터스 컴퓨팅은 언제 어디서나 어떠한 기기를 통해서 자유롭게 컴퓨터에 연결해 각종 정보를 이용할 수 있다. 최근 들어 바코드와 RFID(Radio Frequency IDentification) 기술이 발전하여 고밀도 대량의 정보저장 및 암호인증까지 보유한 특수코드 및 유ㆍ무선 인터넷과 연결시키는 비즈니스모델 등이 등장하면서 신용카드 및 각종 지급결제 서비스 분야 등에 적용되고 있다. 그러나 RFID 칩의 가격이 고가이고 다른 전파의 간섭이 기술적인 문제가 될 수 있다는 점이 상용화가 되는 걸림돌이 되고 있으며, 2차원 바코드 역시 규격화를 이루고 있는 리더기와 소프트웨어 등 대부분 비싼 로열티를 지급하고 들여와야 하는 것이 문제이다. 본 논문은 유비쿼터스 환경에서 2차원 바코드와 RFID를 이용한 사례 및 적용 분야에 대한 최근 기술의 특성을 비교ㆍ분석하여 이에 따른 문제점의 파악 및 활성화 방안에 대해 고찰하고자 한다.

• 한국산업정보학회:학술대회논문집
• /
• 한국산업정보학회 1996년도 추계 학술 발표회 발표논문집
• /
• pp.217-227
• /
• 1996

The new filed of learning control develops controllers that learn to improve their performance at executing a given task , based on experience performing this specific task. In a previous work[6], authors presented a theory of indirect learning control based on use of indirect adaptive control concepts employing simultaneous identification and control. This paper develops improved indirect learning control algorithms, and studies the use of such controller indecentralized systems. The original motivation of the learning control field was learning in robots doing repetitive tasks such as on an asssembly line. This paper starts with decentralized discrete time systems. and progresses to the robot application, modeling the robot as a time varying linear system in the neighborhood of the nominal trajectory, and using the usual robot controllers that are decentralized, treating each link as if it is independent of any coupling with other links. The resultof the paper is to show that stability of the indirect learning controllers for all subsystems when the coupling between subsystems is turned off, assures convergence to zero tracking error of the decentralized indirect learning control of the coupled system, provided that the sample tie in the digital learning controller is sufficiently short.

• 대한화학회지
• /
• 제41권10호
• /
• pp.524-534
• /
• 1997

고무가황체는 고분자인 천연고무 및 합성고무에 여러 가지 유기첨가제를 배합하여 만든다. 이와 같은 배합고무의 분석은 여러 가지 다른 성질을 갖는 다양한 첨가제들로 인하여 많은 전처리 과정을 거쳐야 하므로 그 분석이 용이하지 않다. 이를 극복하기 위하여 열분해 가스크로마토그래피질량분석법(Pyrolysis Gas Chromatography-Mass Spectrometry)을 이용하여 고무가황체 중의 유기 고무약품들의 직접분석을 시도하였다. 점착제인 쿠마론-인덴수지(coumarone-indene resin $C-90^{TM}$), 접착제인 알킬페놀 레조시놀 포름알데히드 수지, 가류지연제 등의 고무첨가제의 열분해과정 및 열분해체들의 질량스펙트럼들을 확인하였고 이를 통하여 복잡한 구성 성분으로 이루어진 가황고무 중에서 이들 배합제의 직접분석이 가능하였다.

• 한국환경농학회지
• /
• 제25권4호
• /
• pp.371-381
• /
• 2006

New proteomic techniques have been pioneered extensively in recent years, enabling the high-throughput and systematic analyses of cellular proteins in combination with bioinformatic tools. Furthermore, the development of such novel proteomic techniques facilitates the elucidation of the functions of proteins under stress or disease conditions, resulting in the discovery of biomarkers for responses to environmental stimuli. The ultimate objective of proteomics is targeted toward the entire proteome of life, subcellular localization biochemical activities, and the regulation thereof. Comprehensive analysis strategies of proteomics can be classified into three categories: (i) protein separation via 2-dimensional gel electrophoresis (2-DE) or liquid chromatography (LC), (ii) protein identification via either Edman sequencing or mass spectrometry (MS), and (iii) proteome quantitation. Currently, MS-based proteomics techniques have shifted from qualitative proteome analysis via 2-DE or 2D-LC coupled with off-line matrix assisted laser desorption ionization (MALDI) and on-line electrospray ionization (ESI) MS, respectively, toward quantitative proteome analysis. In vitro quantitative proteomic techniques include differential gel electrophoresis with fluorescence dyes. protein-labeling tagging with isotope-coded affinity tags, and peptide-labeling tagging with isobaric tags for relative and absolute quantitation. In addition, stable isotope-labeled amino acids can be in vivo labeled into live culture cells via metabolic incorporation. MS-based proteomics techniques extend to the detection of the phosphopeptide mapping of biologically crucial proteins, which ale associated with post-translational modification. These complementary proteomic techniques contribute to our current understanding of the manner in which life responds to differing environment.

• Journal of Plant Biotechnology
• /
• 제37권2호
• /
• pp.196-204
• /
• 2010

With the completion of genome sequencing of several organisms, attention has been focused to determine the function and functional network of proteins by proteome analysis. The recent techniques of proteomics have been advanced quickly so that the high-throughput and systematic analyses of cellular proteins are enabled in combination with bioinformatics tools. Furthermore, the development of proteomic techniques helps to elucidate the functions of proteins under stress or diseased condition, resulting in the discovery of biomarkers responsible for the biological stimuli. Ultimate goal of proteomics orients toward the entire proteome of life, subcellular localization, biochemical activities, and their regulation. Comprehensive analysis strategies of proteomics can be classified as three categories: (i) protein separation by 2-dimensional gel electrophoresis (2-DE) or liquid chromatography (LC), (ii) protein identification by either Edman sequencing or mass spectrometry (MS), and (iii) quanitation of proteome. Currently MS-based proteomics turns shiftly from qualitative proteome analysis by 2-DE or 2D-LC coupled with off-line matrix assisted laser desorption ionization (MALDI) and on-line electrospray ionization (ESI) MS, respectively, to quantitative proteome analysis. Some new techniques which include top-down mass spectrometry and tandem affinity purification have emerged. The in vitro quantitative proteomic techniques include differential gel electrophoresis with fluorescence dyes, protein-labeling tagging with isotope-coded affinity tag, and peptide-labeling tagging with isobaric tags for relative and absolute quantitation. In addition, stable isotope labeled amino acid can be in vivo labeled into live culture cells through metabolic incorporation. MS-based proteomics extends to detect the phosphopeptide mapping of biologically crucial protein known as one of post-translational modification. These complementary proteomic techniques contribute to not only the understanding of basic biological function but also the application to the applied sciences for industry.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.1480-1484
• /
• 2003

This paper presents the design and development of network for housing estate security system. The system can cover up to 961 houses which can be up to 1,200 meters long transfer rate of 9,600 bps. This system uses checking and warning the abnormal situation. More over this system has ability to control switch on/off the electrical equipment in the house via AC line control system. The system consists of 4 parts. The first part is a security system of each house using MCS-51 microcontroller as a central processing unit scan 32 sensors and control 8 appliances and send alarm. The MCS-51 microcontroller received control signal via telephone used DTMF circuit. The second part is distributed two levels master/slave network implementing after RS-485 serial communication standard. The protocol its base on the OSI (Open Systems Interconnection) 7 layers protocol model design focus on speed, reliability and security of data that is transferred. The network security using encrypt by DES algorithm, message sequence, time stamp checking and authentication system when user to access and when connect new device to this system. Flow control in system is Poll/Select and Stop-and-Wait method. The third part is central server that using microcomputer which its main function are storing event data into database and can check history event. The final part is internet system which users can access their own homes via the Internet. This web service is based on a combination of SOAP, HTTP and TCP/IP protocols. Messages are exchanged using XML format [6]. In order to save the number of IP address, the system uses 1 IP address for the whole village in which all homes and appliance in this village are addressed using internal identification numbers. This proposed system gives the data transfer accuracy over 99.8% and maximum polling time is 1,120 ms.

• Korean Chemical Engineering Research
• /
• 제46권6호
• /
• pp.1043-1051
• /
• 2008

폴리올레핀 공정의 melt index(MI or MFI)는 제품의 품질을 결정짓는 가장 중요한 제어변수이다. MI는 실시간으로 측정하는 것이 어렵기 때문에 MI를 예측하여 상관관계를 나타내고자 하는 많은 방법들이 제안되었다. 본 연구에서는 시스템 인식기법을 바탕으로 MI 예측을 위한 새로운 1차의 동적 예측모델을 고안하였다. 이 모델의 예측성능은 등급변경이 수반되는 고밀도 폴리에틸렌 공장의 실제 운전데이터에 근거한 모사로 검증하였으며 다른 예측방법들과의 비교로부터 본 연구에 의한 예측모델의 우수성을 확인하였다. 구성된 MI 동적 예측모델을 토대로 하는 모델예측제어방법의 적용을 통하여 각 단위공정별 MI를 계산하고 운전데이터와 비교하였다. 제어운전의 모사를 통하여 등급변경이 이루어지는 운전 동안의 전이시간과 불량제품 발생량이 현저한 감소를 보임을 확인하였다.

• 융합신호처리학회논문지
• /
• 제2권1호
• /
• pp.44-52
• /
• 2001

본 논문은 수정된 카오틱 신경망(MCNN)을 이용하여 완전 무제약 서체 숫자 인식을 다루고 있다. 카오틱 신경망(CNN)의 동적 특성과 학습과정을 강화함으로써 복잡한 패턴인식 문제를 해결할 수 있는 유용한 신경망으로 수정하였다. MCNN은 신경망 구조와 뉴런 자체가 높은 차수의 비선형 동적특성을 갖고 있으므로 복잡한 서체 숫자를 분류할 수 있는 적합한 신경망이다. 숫자 확인은 원래의 숫자 이미지로부터 특징을 추출하고 MCNN에 근거한 분류기를 이용하여 숫자를 인식한다. MCNN 분류기의 성능은 Canada, Montreal의 Concordia 대학의 숫자 데이터 베이스로 평가하였다. 인식성능의 상대적인 비교를 위해 MCNN 분류기는 리커런트 신경망(RNN) 분류기와 비교하였다. 실험결과에 의하면 인식율은 98.0%이었으며, 이는 MCNN 분류기가 같은 데이터 베이스에 대해 발표되었던 다른 분류기와 RNN 분류기보다 성능이 우수함을 나타낸다.