• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 B
• /
• pp.404-406
• /
• 1998

In this paper, we propose a scaling factor tuning method to improve the performance of fuzzy controller. Tuning rules and reasoning are utilized on-line to determine the scaling factors based on absolute value of the error and its difference. A adaptive evolutionary computation (AEC) is used to search for the optimal tuning rules that will maximize the fitness function. Finally, the proposed fuzzy controller is applied to the angular stabilization of an inverted pendulum.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 B
• /
• pp.897-899
• /
• 1999

This paper presents a scaling factor tuning method to improve the performance of fuzzy logic controller. Tuning rules and reasoning are utilized off-line to determine the scaling factors based on absolute value of the error and its difference. In this paper We proposed a new method to generate fuzzy logic controllers throught genetic algorithm. The developed approach is subsequently applied to the design of proportional plus integral type fuzzy controller for a dc-servo motor control system. The performance of this control system is demonstrated higher than a conventional fuzzy logic controller(FLC).

• 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.

• 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.

• 전자공학회논문지B
• /
• 제33B권3호
• /
• pp.17-27
• /
• 1996

More fuzzy hardware are expected to be utilized in the future to construct fuzzy logic controllers (FLCs). It is hard to find an existing fuzzy hardware which is adopting advanced functions such as self-tuning algorithm in addition to the conventional inference calculation. That is mainly because conventional self-tuning algorithms designed to implement with some hardware circuits is required for fuzzy hardwares to have self-tuning capability. As a first step toward the feature, a novel self-tuning algorithm is proposed in this paper. Based on the search method, the main idea of the proposed algorithm is to detemine valid ranges of input variables of an FLC in order to maximize performance indices fo the control system. The performance indices are so ismple as to be realized by hardware circuit. in dadditon to the conventional scaling-factor adjustment, the algorithm adjusts offset values as well, which, in effect, modifies fuzzy rules of the FLC. To justify the performance of the proposed algorithm, a simulation study is executed.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1997년도 추계학술대회 학술발표 논문집
• /
• pp.50-54
• /
• 1997

In general, the design of fuzzy logic controller has difficulties in the acquisition of expert's knowledge. So, some methods that can optimize the parameters for fuzzy logic controller automatically without expert knowledge was provided. Recently, tuning method for fuzzy logic controller using genetic algorithm(GA) were proposed in many papers. However, those are tuning methods for a part or some part of fuzzy logic controller. In this paper, we proposes auto tuning method for the whole part of tuzzy logic controller, such as parameters of membership functions for antecedence and consequence parts, rule base, scaling factor and the number of rule. Finally, second order dead time plant is provided to show the advantages of the proposed method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1997년도 하계학술대회 논문집 B
• /
• pp.705-707
• /
• 1997

An optimal tuning algorithm of scaling factors is presented in this paper to automatically improve the performance of fuzzy controller. Especially, fuzzy controller has determined an moderate Scaling factor through trial and error. The presented method estimates automatically the optimal values of I/O scaling factors, using modified steepest descent method and this optimal tuning is for nonlinear input/output scaling factors. Simulation results verify the validity of the presented method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 하계학술대회 논문집 B
• /
• pp.717-719
• /
• 1995

In this paper, we analyze the effects of scaling factors on the performance of a fuzzy controller and propose the tuning method for them. The quantitative relation between input and output variables of a fuzzy controller is obtained by using a quasi-linear fuzzy model. An approximate transfer function of a fuzzy controller is derived from the comparison a fuzzy controller with the conventional PID controller. We analyze the effects of scaling factor using this approximate transfer function and propose a fuzzy tuning method based on that of Maeda et al[4].

• 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.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.84-89
• /
• 2003

Fuzzy controller is one of the succeed controller used in the process control in case of model uncertainties. But it my be difficult to fuzzy controller to articulate the accumulated knowledge to encompass all circumstance. Hence, it is essential to provide a tuning capability. There are many parameters in fuzzy controller can be adapted, scale factor tuning of normalized fuzzy controller is one of the adaptation parameter. Two adaptation methods are implemented in this work on an experimental thermal process, which simulate heating process in liquefied petroleum gases (LPG) recovery process in one of petrochemical industries: Gradient decent (GD) adaptation method; supervisory fuzzy controller. A comparison between the two methods is discussed.