• 제목/요약/키워드: 지능기계

검색결과 1,051건 처리시간 0.027초

사례기반 추론에 의한 반도체 패키징 공장의 Cycle-time 예측 모형 개발

  • 김규진;서용무
    • 한국경영정보학회:학술대회논문집
    • /
    • 한국경영정보학회 2007년도 추계학술대회
    • /
    • pp.611-616
    • /
    • 2007
  • 반도체 패키징 공장에서 싸이클타임(Cycle-time)을 정확히 예측하는 것은 납기일 준수를 통해 고객만족도를 향상시킬 수 있고, 보다 효율적인 스케쥴링을 가능하게 하여 공장 가동률을 높일 수 있게 한다. 그러나 반도체 패키징은 제품 종류가 다양하고 제품마다 특화된 기술을 사용할 뿐만 아니라 공정 순서나, WIP에 따라 싸이클타임이 크게 영향을 받아 그 정확한 예측이 매우 어렵기 때문에 현장 전문가의 판단에 의존하는 경우가 많았다. Fab공정의 경우 전문가를 도와 좀 더 정확한 예측에 도움을 주기 위해 그 동안 전통적 통계 기법 및 시뮬레이션에 기반한 의사결정 모형이 많이 연구되었는데, 최근에는 기계학습 및 인공지능 기법을 사용한 연구가 눈에 띄고 있으며 기존의 방법보다 우수한 성능을 보여 주는 것으로 나타났다. 하지만 아직 기계학습 및 인공지능을 이용한 충분한 연구가 진행되지 못하고 있는 실정이다. 따라서 본 연구에서는 사례기반 추론을 사용하여 패키징 공정의 싸이클타임을 예측하고자 하였으며 그 성능을 인공신경망 모형, 의사결정나무 모형, 그리고 해당 분야 전문가의 예측치와 비교하였다. 실험결과에 따르면 사례기반추론 모형이 가장 뛰어난 성능을 보이는 것으로 나타났다.

  • PDF

지능형 공작기계 설계 지원 시스템 개발 (Development of Intellingent Design Support System for Machine Tools)

  • 차주헌;김종호;박면웅;박지형
    • 한국정밀공학회:학술대회논문집
    • /
    • 한국정밀공학회 1995년도 추계학술대회 논문집
    • /
    • pp.1022-1027
    • /
    • 1995
  • We present a framework of an intelligent design support system for embodiment design of machine tools which can support efficiently and systematically the machine design by utilizing design knowledge such as objects(part), know-how, public, evaluation, and procedures. The design knowledge of machining center has been accumulated through interview with design experts of machine tool companies. The processes of embodiment design of machining center are established. We also introduce a hybrid knowledge representation so that the systm can easily deal with various and complicated design knowledge. The intelligent design system is being developed on the basis of object-oriented programming, and all parts of a design object, machining center, are also classified by the object-oriented modeling. For the demonstration of effectiveness of the suggested system, a structural design system for machine tools is implemented.

  • PDF

형상기억합금을 이용한 지능형 고분자 복합재료의 설계 (Design of an Intelligent Polymer-Matrix-Composite Using Shape Memory Alloy)

  • 정태헌
    • 대한기계학회논문집A
    • /
    • 제21권10호
    • /
    • pp.1609-1618
    • /
    • 1997
  • Thermo-mechanical behaviors of polymer matrix composite(PMC) with continuous TiNi fiber are studied using theoretical analysis with 1-D analytical model and numerical analysis with 2-D multi-fiber finite element(FE) model. It is found that both compressive stress in matrix and tensile stress in TiNi fiber are the source of strengthening mechanisms and thermo-mechanical coupling. Thermal expansion of continuous TiNi fiber reinforced PMC has been compared with various mechanical behaviors as a function of fiber volume fraction, degree of pre-strain and modulus ratio between TiNi fiber and polymer matrix. Based on the concept of so-called shape memory composite(SMC) with a permanent shape memory effect, the critical modulus ratio is determined to obtain a smart composite with no or minimum thermal deformation. The critical modulus ratio should be a major factor for design and manufacturing of SMC.

지능생산시스템을 위한 PC 기반 개방형 머신제어기 (PC based Open Architecture Machine Controller for Intelligent Manufacturing system)

  • 박경수;최경현;이석희
    • 대한기계학회:학술대회논문집
    • /
    • 대한기계학회 2000년도 춘계학술대회논문집A
    • /
    • pp.884-889
    • /
    • 2000
  • This paper describes the design and implementation of PCOAMC(PC-based Open Architecture Machine Controller) to be flexible and independent from the vendor-oriented hardware and software structure. This openness approach is able to enhance an intelligence and integration of a manufacturing system. The development methodology of PcOAMC is an Object-Oriented approach, and all modules in PcOAMC are modelled using UML(Unified Modelling Language) that provides an easy understanding and modification. In order to demonstrate an applicability of PcOAMC, a simple test has been executed by using the Client-Server system consisting of two PcOAMCS and bus monitoring system. The good results have been obtained, so that the developed controller is expected to be embedded into IMS(Intelligent Manufacturing System) as a basic unit.

  • PDF

지능형 순항 제어 시스템 모델링 및 제어 (A Modeling and Control of Intelligent Cruise Control Systems)

  • 이세진;홍진호;이경수
    • 대한기계학회논문집A
    • /
    • 제25권2호
    • /
    • pp.283-288
    • /
    • 2001
  • A throttle/brake control law for the intelligent cruise control(ICC) systems has been proposed in this paper. The ICC system consists of a vehicle detection sensor, the control algorithm and a throttle/brake actuators. For the control of a throttle/brake system, we introduced a solenoid-valve-controlled electronic vacuum booster and a step-motor controlled throttle actuator. Nonlinear computer model for the electronic vacuum booster has been developed and the simulations were per formed using a complete nonlinear vehicle model. The results indicate the proposed throttle/brake control law can provide the ICC system with an optimized performance.

실험법에 의한 지능성 복합체의 회복응력 측정에 관한 연구 (A Study on the Measurement for the Recovery Stress of Intelligent Composite by Experiment)

  • 황재석;이효재
    • 대한기계학회논문집A
    • /
    • 제27권4호
    • /
    • pp.515-523
    • /
    • 2003
  • Shape memory is physical phenomenon which a platically metal is restored to its original shape by a solid state phase change by heating. TiNi alloy the most effective material in the shape memory alloy(SMA). To study(measure) recovery stress of intelligent composite. Ti50-Ni50 shape memory matrix with prestrain SMA fiber. When SMA fiber of the intelligent composite is heated over austenite starting temperature(As) by electric heating. a recovery stress are generated. The recovery stress of the intelligent composite was measured by strain gage or photoelastic experiment. Measuring method of recovery stress by photoelastic experiment was developed in this research. It was certified that photoelastic experiment was more effective and more precise than strain gage method in the measurement of recovery stress.

효과적인 필기체 인식을 위한 인공지능 알고리즘 (Artificial Intelligence Algorithms for Identification of Handwriting)

  • 김승주;이재영;한종기
    • 한국방송∙미디어공학회:학술대회논문집
    • /
    • 한국방송∙미디어공학회 2016년도 추계학술대회
    • /
    • pp.151-153
    • /
    • 2016
  • 최근 스마트폰, PC, 태블릿 같은 전자기기들이 발전하면서 기계를 통해 소통하는 시대가 왔다. 기계와 소통하기 위해 우리가 사용하는 문자를 인식하는 것은 중요한 일이다. 이런 전자기기들이 문자, 영상인식을 해야 할 필요성이 더욱 증가함에 따라 머신러닝의 중요성이 대두되었다. 머신러닝은 컴퓨터의 학습을 위해 알고리즘과 기술을 개발하는 분야를 말한다. 머신러닝의 기법과 관련된 알고리즘의 종류는 수없이 많다. 그 중에서도 Neural Network는 사람의 뇌 신경구조를 토대로 착안하여 네트워크를 만들고 이를 학습에 이용한 머신러닝 기법이다. 이런 인공지능 알고리즘인 Neural Network 구조를 바탕으로 특징을 추출하여 학습을 하는 Convolution Neural Network 기법의 사용이 늘고 있다. 본 논문에서는 Neural Network와 Convolution Neural Network의 알고리즘을 이용한 필기체 인식 실험을 하고 그 내용을 비교하였다.

  • PDF

웨이퍼 연마용 지능형 연삭시스템 개발 (Development of the intelligent grinding system for wafer grinding)

  • 김동석;최춘규;하상백;이상직
    • 대한기계학회:학술대회논문집
    • /
    • 대한기계학회 2004년도 춘계학술대회
    • /
    • pp.1082-1086
    • /
    • 2004
  • In silicon wafer manufacturing process, the grinding process has been adopted to improve the flatness of wafer. The grinding of wafer is usually used by the infeed grinding machine. The infeed grinding machine has been depended on imports. Therefore, it is necessary to develop the infeed grinding machine because the demand of the infeed grinding machine is increasing more and more. This paper describes the technologies of infeed grinding machine and intend to introduce the studies in the development of the intelligent grinding system for grinding of wafer. The air bearing spindle for the infeed grinding machine was developed by domestic technologies and the grinding part design of the intelligent grinding system for wafer grinding was completed.

  • PDF

비구면 가공용 지능형 연삭 시스템 개발 (Development of intelligent grinding system for aspherical surface machining)

  • 백승엽;이해동;김성철;이은상
    • 대한기계학회:학술대회논문집
    • /
    • 대한기계학회 2004년도 춘계학술대회
    • /
    • pp.1099-1104
    • /
    • 2004
  • As consumer in optics, electronics, aerospace and electronics industry grow, the demand for ultra precision aspherical surface lens increases higher. To enhance the precision and productivity of ultra precision aspherical surface micro lens, the following specification of ultra precision grinding system is required: the highest rotational speed of the grinder is 100,000rpm and its turning accuracy is $0.1{\mu}m$, positioning accuracy is $0.1{\mu}m$. The development process of the grinding system for the ultra precision aspherical surface micro lens for optoelectronics industry is introduced. In the work reported in this paper, an intelligent grinding system for ultra precision aspherical surface machining was designed by considering the factors affecting the surface roughness and profiles accuracy. An aerostatic form was adopted to build the spindle of the workpiece and the spindle of grinder and ultra precision LM guide way was adopted in this system.

  • PDF

미지물체를 안전하게 잡기 위한 지능형 그리퍼의 제어장치 개발 (Development of Intelligent Gripper Control Device to Safely Grip Unknown Objects)

  • 김한솔;김갑순
    • 한국기계가공학회지
    • /
    • 제21권4호
    • /
    • pp.31-38
    • /
    • 2022
  • In this study, we designed and manufactured an intelligent gripper-control device to safely grip unknown objects. The gripper control device consists of a DSP circuit, power supply circuit, communication circuit, and amplifier circuit diagrams. The DSP is used because the values of the 3-axis force sensor to which the gripper is attached are measured and calculated at high speeds. The gripping force is determined based on this value, and the object must be safely gripped with the determined value. A basic characteristic test of the control device of the manufactured intelligent gripper was conducted, and it was confirmed that it operated safely.