• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2006년도 춘계학술대회 논문집
• /
• pp.271-272
• /
• 2006

This paper presented a new experimental gain tuning technique for a Multi-Axis Servo System. First, the investigation for proportional gain of velocity control loop by using a Dynamic Signal Analyzer (DSA) was performed. Using the FUNCTION characteristic of DSA based on the Bode plot, the Bode plot of open loop transfer function was obtained. In turn, the integral gain of a servo controller can be found out by using the Integration time constant extracted from the Bode plot of open loop transfer function. In the meanwhile, the positional gain of the servo controller has been obtained by using the Bode plot of the closed loop transfer function. We have also proposed the technique to find out an optimal parameter of a notch filter, which has a great influence on vibration reduction, by using the damping factor extracted from the Bode plot of closed loop transfer function.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2008년도 춘계종합학술대회 A
• /
• pp.213-217
• /
• 2008

본 논문에서는 표준 IMC-PID 제어기에 시스템 식별상의 위상 조절 인자를 설계 변수로 추가함으로 새로운 IMC-PID 제어기 설계 방법을 제안한다. 제안된 제어기의 동조법으로 해석적으로 구해진 주파수 영역 특성을 통해 주파수 영역 성능을 보장하도록 하는 설계 변수의 구속 조건 하에서 시간영역의 적분 오차 지표를 가격함수로 최소화하는 최적 해를 얻는다. 이 제안된 동조법은 다양한 설계 사양을 동시에 고려하도록 하여 기존의 IMC-PID의 단일 설계 변수 동조가 갖는 비유연성 문제를 개선하였다. 사례 연구에서 기존의 동조법들과의 성능 비교를 통해 그 유용성을 검토했다.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회A
• /
• pp.1857-1862
• /
• 2008

Resonant frequency tuning of micro devices is essential to achieve performance uniformity and high sensitivity. Previously reported frequency tuning methods using electrostatic force or mass deposition are not directly applicable to non-conducting polymer devices and have limitations such as dielectric breakdown or low tunable bandwidth. In this paper, thermally frequency-tunable microactuators with poly-dimethylsiloxane membranes are proposed. Permanent and/or nonpermanent frequency tunings are possible using a simple temperature control of the device. Resonant frequency and Q-factor variations of devices according to temperature change were studied using a micro heater and laser Doppler vibrometer. The initial resonant frequencies determined by polymer curing and hardening temperatures are reversibly tuned by thermal cycles. The measured resonant frequency of 9.7 kHz was tuned up by ${\sim}25%$ and Q-factor was increased from 14.5 to 27 as the micro heater voltage increased from 0 to 70 V.

• 응용통계연구
• /
• 제24권6호
• /
• pp.1225-1235
• /
• 2011

본 논문은 수치기상모형에 의해 계산된 수치기상모의풍속($1km{\times}1km$ 해상도)의 정확도를 향상시키기 위한 통계적 보정법을 제안하였다. 이를 위해 남한전역을 $1km{\times}1km$ 격자로 나눈 지점(345,682지점)에 적합한 통계적 바람장 모형으로 부터 남한지역의 바람장을 추정하는 절차와 격자지점별/월별 보정인자를 계산하여 추정된 바람장과 수치 기상모의풍속간의 간극을 보정하는 절차로 이루어진 보정인자법을 개발하였다. 또한 75개 기상관측소지점에서 계산된 수치기상모의풍속자료에 보정인자법을 적용시켜 본 논문에서 제안된 보정법의 유용성을 보였다.

• 한국IT서비스학회지
• /
• 제19권5호
• /
• pp.83-91
• /
• 2020

Language Models such as BERT has been an important factor of deep learning-based natural language processing. Pre-training the transformer-based language models would be computationally expensive since they are consist of deep and broad architecture and layers using an attention mechanism and also require huge amount of data to train. Hence, it became mandatory to do fine-tuning large pre-trained language models which are trained by Google or some companies can afford the resources and cost. There are various techniques for fine tuning the language models and this paper examines three techniques, which are data augmentation, tuning the hyper paramters and partly re-constructing the neural networks. For data augmentation, we use no-answer augmentation and back-translation method. Also, some useful combinations of hyper parameters are observed by conducting a number of experiments. Finally, we have GRU, LSTM networks to boost our model performance with adding those networks to BERT pre-trained model. We do fine-tuning the pre-trained korean-based language model through the methods mentioned above and push the F1 score from baseline up to 89.66. Moreover, some failure attempts give us important lessons and tell us the further direction in a good way.

• 한국산업융합학회 논문집
• /
• 제3권1호
• /
• pp.3-8
• /
• 2000

This paper describes a development of self tuning scheme for PI and PO type fuzzy controllers. The output scaling factor(SF) is adjusted on-line by fuzzy rules according to the current trend of the controlled process. The rule-base for tuning the output SF is defined on error and change of error for the controlled variable using the most natural and unbiased membership functions. Simulation results demonstrate the better control performance can be achieved in comparison with Ziegler-Nichols(Z-N) PID controllers.

• 한국철도학회:학술대회논문집
• /
• 한국철도학회 2010년도 춘계학술대회 논문집
• /
• pp.514-520
• /
• 2010

This paper is about control algorithms that bi-direction DC-DC Converter using Super Capacitor and regenerative power from DC feeding system in train. In order to take advantage of regenerative energy efficient, charge and discharge level value of energy storage system serve as an important factor. Respect to output fluctuations of the substation and catenary voltage changing, we offers Charge-Discharge Auto Level Tuning Algorithms to improve system following of Energy Storage System.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 춘계 학술대회 학술발표 논문집
• /
• pp.325-329
• /
• 2003

This paper discuss the method of the system's efficient control using a Intelligent hybrid algorithm in nonlinear dynamics systems. Existing neural network and genetic algorithm for the control of non-linear systems work well in static states. but it be not particularly good in changeable states and must re-learn for the control of the system in the changed state. This time spend a lot of time. For the solution of this problem we suggest the intelligent hybrid self-tuning controller. it includes neural network, genetic algorithm and immune system. it is based on neural network, and immune system and genetic algorithm are added against a changed factor. We will call a change factor an antigen. When an antigen broke out, immune system come into action and genetic algorithm search an antibody. So the system is controled more stably and rapidly. Moreover, The Genetic algorithm use the memory address of the immune bank as a genetic factor. So it brings an advantage which the realization of a hardware easy.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1994년도 추계학술대회 논문집 학회본부
• /
• pp.338-341
• /
• 1994

This paper presents a Fuzzy-PID controller based on Fuzzy logic. Up to now PID controller has had the difficulty of obtaining the optimal gain, and Fuzzy controller has had the difficulty of determining scale factor affecting the performance of control. So that a Fuzzy-PID controller is presented here self tuning of the scale factor and optimal gain. The results of simulation show a good performance in comparison with Ziegler-Nichols controller, having the generality of determining the components of scale factor in Fuzzy rule.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1992년도 한국자동제어학술회의논문집(국제학술편); KOEX, Seoul; 19-21 Oct. 1992
• /
• pp.265-271
• /
• 1992

It is the purpose of this paper to present a dialogical designing method for control system using a rough grasp of the unknown process property. We deal with a single-input single-output feedback control system with a fuzzy controller. The process property is roughly estimated by the step response, and the fuzzy controller is interactively modified according to the operator's requests. The modifying rules mainly derived from computer simulation are useful for almost every process, such as an unstable process and a non-minimum phase process. The fuzzy controller is tuned by taking notice of four characteristics of the step response: (1) rising time, (2) overshoot, (3) amplitude and (4) period of vibration. The tuning position of the controller is fourfold: (1) antecedent gain factor GE or GCE, (2) consequent gain factor GDU, (3) arrangement of the antecedent fuzzy labels and (4) arrangement of the control rules. The rules give an instance to the respective items of the controller in an effective order. The modified fuzzy PI controller realizes a good response of a stable process. However, because the GDU tuning becomes difficult for the unstable process, it is necessary to evaluate the stability of the process from the initial step response. The fuzzy PI controller is applied to the process whose initial step response converges with GDU tuning. The fuzzy PI controller with modified sampling time is applied to the process whose step response converges under the repeated application of the GDU tuning. The fuzzy PD controller is applied to the process whose step response never converges by the GDU tuning.