• Nuclear Engineering and Technology
• /
• 제27권3호
• /
• pp.317-326
• /
• 1995

증기발생기의 수위제어시스템에 대해 특히 저출력시 수위제어의 문제점을 분석고찰하고 퍼지제어기 법을 기반으로 한 안정되고도 신속한 수위제어에 관한 연구가 주로 수행되었다. 문제해결의 한 방안으로서, 중요 제어변수임에도 불구하고 저출력운전시 저유량구간의 추정불량으로 인해 사용할 수 없는 유량신호를 대신하여 밸브개도를 이용한 대체정보를 채용하였으며 또한 소속함수크기의 유동적인 조정방법을 이용하여 수위오차가 크게 발생한 과도상태시에는 신속한 수위회복이 이루어지도록 하였다. 실제운전환경에서 제시된 제어기를 튜닝 하기 위한 방법으로서 제어성능지수 및 decent method를 이용한 소속함수의 self-tuning 기법을 제시하였다. 원자력 연구소의 연수원에 설치된 교육훈련용 시뮬레이터에서 수행된 실험결과는 제시된 제어기 및 튜닝방법의 안정되고 우수한 성능구현 및 실질적 인 유용성을 보여주고 있다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 전문대학교육위원
• /
• pp.155-157
• /
• 2005

This paper shows how to design a MPPT control of PV system using neural network PI self tuning. The conventional self-tuning methods have the voltage control problem of nonlinear PV system which can't adapt against any kinds of noise or operation circumstances. In this paper, supposed to solve these problem to PI parameters controller algorithm using ANN. In the proposed algorithm, the parameters of the controller were adjusted to reduce by on-line system the error of the output voltage of DC-DC chopper. In this process, EBPA NN was constituted to an output error value of a DC-DC chopper and conspired an input and output. The performance of the self-tuning controller is compared with that of the PI controller tuned by conventional method. The effectiveness of the proposed control method is verified thought the Matlab Simulink.

• 전기학회논문지
• /
• 제63권4호
• /
• pp.509-518
• /
• 2014

Though various techniques have been studied as a way of adjusting parameters of PID controllers, no perfect method of determining parameters is available to date. Especially the deign of PID controller for unstable processes with dead time(UPWDT) is even more difficult due to various reasons. Generally the existing design procedures for UPWDT involve deriving formulas to meet gain and phase margin specifications, or using inner loop to stabilize UPWDT before applying PID controller. In this paper, the dual-input describing function(DIDF) method is proposed, by which the performance and robustness of the closed-loop system can be improved. The method is based on moving the critical point (-1+j0) of Nyquist stability to a new position arbitrarily selected on the complex plane. This can be done by determining appropriate coefficients of the DIDF. As a result, we can easily determine parameters of PID-type controller by using existing conventional tuning methods for stable or unstable systems. Simulation results are included to show the effectiveness of the proposed method.

• ETRI Journal
• /
• 제40권4호
• /
• pp.421-434
• /
• 2018

The application of deep neural networks (DNNs) to connect the world with cyber physical systems (CPSs) has attracted much attention. However, DNNs require a large amount of memory and computational cost, which hinders their use in the relatively low-end smart devices that are widely used in CPSs. In this paper, we aim to determine whether DNNs can be efficiently deployed and operated in low-end smart devices. To do this, we develop a method to reduce the memory requirement of DNNs and increase the inference speed, while maintaining the performance (for example, accuracy) close to the original level. The parameters of DNNs are decomposed using a hybrid of canonical polyadic-singular value decomposition, approximated using a tensor power method, and fine-tuned by performing iterative one-shot hybrid fine-tuning to recover from a decreased accuracy. In this study, we evaluate our method on frequently used networks. We also present results from extensive experiments on the effects of several fine-tuning methods, the importance of iterative fine-tuning, and decomposition techniques. We demonstrate the effectiveness of the proposed method by deploying compressed networks in smartphones.

• 전자공학회논문지CI
• /
• 제39권4호
• /
• pp.19-27
• /
• 2002

역전파 알고리즘의 성능 개선을 위해서 학습율을 자동 조정하는 방식을 제안하였다. 제안한 방식은 각각의 연결강도의 학습율을 퍼지 논리 시스템을 이용하여 자동 조정하는 방식으로 각각의 연결강도에 대해서 ${\Delta}$와 $\bar{{\Delta}}$를 구하여 퍼지 논리 시스템의 입력으로 사용하고, 학습율을 출력으로 사용하였다. 제안한 방식을 N-패리티 문제, 함수 근사, 숫자 패턴 분류에 대한 시뮬레이션 결과 일반적인 역전파 알고리즘, 모멘텀 방식, Jacobs의 delta-bar-delta 방식보다 성능이 개선됨을 확인하였다.

• Journal of Electrical Engineering and Technology
• /
• 제12권1호
• /
• pp.72-77
• /
• 2017

This paper proposes the explicit tuning rule of multi-loop PID controller for brushless direct current motors to predict the system behaviors in time and frequency domains, using properties of the convex combination method. The convex set of the proposed controllers formulates the envelope to satisfy the performances in time and frequency domains. The final control parameters are determined by solving the convex optimization problem subject to the constraints which are represented as convex set of time domain performances. The effectiveness of the proposed control method is shown in the numerical simulation, in which controller tuning algorithm and dynamics of brushless DC motor are well taken into account.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 하계학술대회 논문집 B
• /
• pp.852-854
• /
• 1995

This paper proposes a new identification method that utilizes fuzzy inference in parameter identification. The prosed system has an additional control loop where a real plant has replaced by a plant model. Fuzzy rules describe the relationship between comparison results of the features and magnitude of modification in the model parameter values. In this paper, the tuning method which determines parameters of PID controller automatically is described through applying this algorithm to DC servo motor. And we intend to investigate effectiveness of the method by experiments. This method is effective in auto-tuning because the response of the closed loop has verified. The simulated and the experimental results of the dc servo motor are shown to confirm the viability of this method.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2001년도 Proceedings ICPE 01 2001 International Conference on Power Electronics
• /
• pp.462-465
• /
• 2001

The wound induction motor can provide high starting torque and reduced starting current simultaneously by inserting large scale resistor. And this technique is one of the well known methods among the induction motor starting methods and generally used for heavy load starting such as Crain and Cement factories. The conventional PI controller has been widely used in industrial application due to the simple control algorithm and in general, PI controller is used for control of current, torque, position, and speed for the wound induction motor drive system. However, the system may result in poor performance since sensors have to be used, which in turn is limited by the environmental condition. Recently, to overcome these problems, many sensorless vector control methods for the wound induction motor have been studied. This paper presents MRAS method with on-line stator resistance tuning for sensorless vector control of the wound induction motor drive. In conventional MRAS method, in low frequency, stator resistance variation can result in poor performance. Therefore, to overcome several shortages of the conventional MRAS caused by parameter variation and enhance robustness of the sensor less vector control, this paper investigates a MRAS method with on-line stator resistance tuning for sensorless vector control of the wound induction motor. The validity and effectiveness of the proposed method is verified through digital simulation.

• 대한전기학회논문지
• /
• 제41권9호
• /
• pp.1071-1084
• /
• 1992

In this paper, the method for navigation and obstacle avoidance of the autonomous mobile robot is proposed. The proposed algorithms are based on the fuzzy inference system which is able to deal with imprecise and uncertain information. The self-tuning algorithm, which adopts the simplex method, modifies the parameters of membership functions of the input-output linguistic variables by changing the support of these fuzzy sets according to the integral of absolute error(IAE) of the system response. The wall-follwing navigation and obstacle avoidance of the mobile robot are based on range data measured from the internal sensors(encoder) and the outer sensors(sonar sensor). In addition, the algorithm for the obstacle detection proposed in this paper is based on the expert's experience. Finally, the effectiveness of navigation and obstacle avoidance algorithm is demonstrated through simulation and experiment.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 D
• /
• pp.2079-2081
• /
• 2002

R/C helicopter has been used to several fields of military affairs, investigation searching and toys because it has small size hovering and vortical take-off characteristics etc. Therefore it needs more realizable control method. The paper introduces simulation and experimental results for control of a helicopter training simulator by self tuning control method. To realize the disturbance rejection and the given reference tracking, a least common multiple polynomial between the reference and disturbance model polynomials is operated to the plant model. The effectiveness of the method is shown by simulation and experimental results for a helicopter training simulator with two degree of freedom.