• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.238-238
• /
• 2000

In this paper, to increase the utilization of uranium resources contained in the spent fuel, the spent fuel is reused. for this, the spent fuel is dismantled or spent fuel rod is extracted from the spent fuel assembly. Therefore, to achieve the performance of compacting the spent fuel assembly, we proposed the controller consisting of adaptive and fuzzy with teaming algorithm. In order to show the performance of proposed algorithm compares, we compared the controller with conventional controller in plant.

• 대한임베디드공학회논문지
• /
• 제3권2호
• /
• pp.119-125
• /
• 2008

This paper presents the polygon-based Q-leaning and Cascade Support Vector Machine algorithm for object search with multiple robots. We organized an experimental environment with ten mobile robots, twenty five obstacles, and an object, and then we sent the robots to a hallway, where some obstacles were lying about, to search for a hidden object. In experiment, we used four different control methods: a random search, a fusion model with Distance-based action making (DBAM) and Area-based action making (ABAM) process to determine the next action of the robots, and hexagon-based Q-learning and dodecagon-based Q-learning and Cascade SVM to enhance the fusion model with DBAM and ABAM process.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 1998년도 전력전자학술대회 논문집
• /
• pp.384-387
• /
• 1998

This paper presents a newly developed speed sensorless drive using RLS(Recursive Least Squares) based on Neural Network Training Algorithm. The proposed algorithm based on the RLS has just the time-varying learning rate, while the well-known back-propagation (or generalized delta rule) algorithm based on gradient descent has a constant learning rate. The number of iterations required by the new algorithm to converge is less than that of the back-propagation algorithm. The RLS based on NN is used to adjust the motor speed so that the neural model output follows the desired trajectory. This mechanism forces the estimated speed to follow precisely the actual motor speed. In this paper, a flux estimation strategy using filter concept is discussed. The theoretical analysis and experimental results to verify the effectiveness of the proposed analysis and the proposed control strategy are described.

• Journal of Biosystems Engineering
• /
• 제31권3호
• /
• pp.203-208
• /
• 2006

This study was conducted to improve the control performance of a current variable-rate controller for granular fertilizers. Simulation model was developed. Optimized proportional, integral and derivative gains were determined by simulation model using 2nd order PID gain learning algorithm, and these control gains were evaluated through the field tests. Important results of this study are as follows; 1. Principles of pre-existing variable-rate application of granular fertilizers were investigated. 2. Simulation model of a PID controller that could simulate the control system was developed by using Matlab/Simulink program. The program was to determine PID control coefficients through the simulation model and 2nd order PID gain learning algorithm. 3. PID control coefficients obtained from the simulation were applied to the developed model. When the step input was given, Maximum overshoot were 1.96%, rise time were 0.05 sec, settling time were 0.06 sec and steady state error were 0.21 % respectively. 4. The simulation model was verified through field tests. The errors of maximum overshoot were 10%, rise time were 0.11 sec, settling time were 0.40 sec and steady state error were 8% because of loads and noises. Rise time was decreased to one third of that of the pre-existing system. 5. If the speed of a fertilizing machine is $0.3{\sim}0.6\;m/s$ and the maximum rotation speed of a discharging roller is 64 rpm, rise time would be 0.26 sec and fertilizing machine would cover the distance of $0.07{\sim}0.15\;m$ with settling time of 0.4 sec, fertilizing machine would cover the distance of $0.12{\sim}0.24\;m$.

• 한국수자원학회논문집
• /
• 제54권spc1호
• /
• pp.1071-1081
• /
• 2021

본 연구에서는 기상청에서 수행하는 기존의 기상 관측에 대한 품질관리 절차 이외에 향후 스마트시티 등에서 활용될 수 있는 머신러닝 기반의 Internet of Things (IoT) 도시기상 관측 자료에 대한 품질검사 기준을 제안한다. 현재 기상청에서 종관기상관측(Automated Synoptic Observing System, ASOS)과 방재기상관측(Automatic Weather System, AWS) 기반으로 설정한 기준이 도시기상에 적합한지 확인하기 위하여 서울시에 설치된 SKT AWS 자료를 기반으로 사용성을 검증하였고, IoT 자체의 데이터가 가지는 특성을 고려하여 최종적으로 머신러닝 기반의 품질검사 알고리즘을 제안하였다. 품질검사 방법으로는 IoT 기기 자체에 대한 결측값 검사, 값 패턴 검사, 충분 데이터 검사, 통계적 범위 이상 검사, 시간값 이상 검사, 공간값 이상 검사를 먼저 수행하고, 기상청에서 제시하고 있는 기상 관측에 대한 품질검사인 물리한계검사, 단계검사, 지속성 검사, 기후범위 검사, 내적 일치성 검사를 5가지 기상요소에 대하여 각각 수행하였다. 제안한 알고리즘의 검증을 위하여 인천광역시 송도에 위치한 관측소에 실제 IoT 도시기상관측 데이터에 이를 적용하였다. 이를 통해 기존의 기상청 QC로는 확인할 수 없었던 IoT 기기가 가질 수 있는 결함을 확인할 수 있고, 알고리즘에 대한 검증을 진행하여 향후 스마트시티에 설치될 IoT 기상관측기기에 대한 품질검사 방법을 제안한다.

• 제어로봇시스템학회논문지
• /
• 제5권1호
• /
• pp.95-104
• /
• 1999

In this paper we propose a fast and precise control algorithm for a mobile robot, which aims at the self-tuning control applying two multi-layered neural networks to the structure of computed torque method. Through this algorithm, the nonlinear terms of external disturbance caused by variable task environments and dynamic model errors are estimated and compensated in real time by a long term neural network which has long learning period to extract the non-linearity globally. A short term neural network which has short teaming period is also used for determining optimal gains of PID compensator in order to come over the high frequency disturbance which is not known a priori, as well as to maintain the stability. To justify the global effectiveness of this algorithm where each of the long term and short term neural networks has its own functions, simulations are peformed. This algorithm can also be utilized to come over the serious shortcoming of neural networks, i.e., inefficiency in real time.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 B
• /
• pp.942-944
• /
• 1999

This paper presents a genetic learning algorithm with partial update technique in application to active control system. Proposed algorithm divides active control system into two parts, real time control part and control parameter update part. This genetic algorithm has global convergent advantage and is expected to be applied easily to real time active noise and vibration control systems. Computer simulation was performed.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2005년도 춘계학술대회 학술발표 논문집 제15권 제1호
• /
• pp.217-220
• /
• 2005

The RPO algorithm is a recently developed tool in the area of reinforcement Loaming, And it has been shown In be very successful in several application problems. In this paper, we consider a robot-control problem utilizing a modified RPO algorithm in which its critic network is adapted via RLS(Recursive Least Square) algorithm. We also developed a MATLAB-based animation program, by which the effectiveness of the training algorithms were observed.

• International Journal of Safety
• /
• 제1권1호
• /
• pp.1-6
• /
• 2002

It is important to control the high accurate position and force to prevent unexpected accidents by a robot manipulator. Direct-drive robots are suitable to the position and force control with high accuracy, but it is difficult to design a controller because of the system's nonlinearity and link-interactions. This paper is concerned with the study of the stabilization force control of direct-drive robots. The proposed algorithm is consists of the feedback controllers and the neural networks. After the completion of learning, the outputs of feedback controllers are nearly equal to zero, and the neural networks play an important role in the control system. Therefore, the optimum adjustment of control parameters is unnecessary. In other words, the proposed algorithm does not need any knowledge of the controlled system in advance. The effectiveness of the proposed algorithm is demonstrated by the experiment on the force control of a parallelogram link-type robot.

• 전기전자학회논문지
• /
• 제21권2호
• /
• pp.136-141
• /
• 2017

본 논문에서는 딥러닝을 PC에 적용하기 위한 메모리 최적화에 관한 알고리즘을 제안한다. 제안된 알고리즘은 일반 PC에서 기존의 딥러닝 구조에서 요구되는 연산처리 과정과 데이터 량을 감소시켜 메모리 및 연산처리 시간을 최소화한다. 본 논문에서 제안하는 알고리즘은 분별력이 있는 랜덤 필터를 이용한 컨볼루션 층 구성 과정, PCA를 이용한 데이터 축소 과정, SVM을 사용한 CNN 구조 생성 등의 3과정으로 이루어진다. 분별력이 있는 랜덤 필터를 이용한 컨볼루션 층 구성 과정에서는 학습과정이 필요치 않아서 전체적인 딥러닝의 학습시간을 단축시킨다. PCA를 이용한 데이터 축소 과정에서는 메모리량과 연산처리량을 감소시킨다. SVM을 사용한 CNN 구조 생성에서는 필요로 하는 메모리량과 연산 처리량의 감소 효과를 극대화 시킨다. 제안된 알고리즘의 성능을 평가하기 위하여 예일 대학교의 Extended Yale B 얼굴 데이터베이스를 사용하여 실험한 결과, 본 논문에서 제안하는 알고리즘이 기존의 CNN 알고리즘과 비교하여 비슷한 성능의 인식률을 보이면서 연산 소요시간과 메모리 점유율에 있어 우수함이 확인되었다. 본 논문에서 제안한 알고리즘을 바탕으로 하여 일반 PC에서도 많은 데이터와 연산처리를 가진 딥러닝 알고리즘을 구현할 수 있으리라 기대된다.