• 한국지능시스템학회논문지
• /
• 제11권3호
• /
• pp.231-240
• /
• 2001

Conventional expert systems has been criticized due to its lack of capability to adapt to the changing decision-making environments. In literature, many methods have been proposed to make expert systems more environment-adaptive by incorporating fuzzy logic and neural networks. The objective of this paper is to propose a new approach to building a self-evolving expert system inference mechanism by integrating fuzzy neural network and fuzzy rule extraction technique. The main recipe of our proposed approach is to fuzzify the training data, train them by a fuzzy neural network, extract a set of fuzzy rules from the trained network, organize a knowledge base, and refine the fuzzy rules by applying a pruning algorithm when the decision-making environments are detected to be changed significantly. To prove the validity, we tested our proposed self-evolving expert systems inference mechanism by using the bankruptcy data, and compared its results with the conventional neural network. Non-parametric statistical analysis of the experimental results showed that our proposed approach is valid significantly.

• Journal of Advanced Marine Engineering and Technology
• /
• 제23권4호
• /
• pp.480-487
• /
• 1999

In this paper we propose a neural network precompensated PID(NNP PID) controller for load frequency control of 2-area power system. While proportional integral derivative(PID) controllers are used in power system they have many problems because of high nonlinearities of the power system So a neural network-based precompensation scheme is adopted into a conventional PID controller to obtain a robust control to the nonlinearities. The applied neural network precompen-sator uses an error back-propagation learning algorithm having error and change of error as inputand considers the changing component of forward term of weighting factor for reducing of learning time. Simulation results show that the proposed control technique is superior to a conventional PID controller and an optimal controller in dynamic responses about load disturbances. The pro-posed technique can be easily implemented by adding a neural network precompensator to an existing PID controller.

• 대한기계학회논문집
• /
• 제16권7호
• /
• pp.1322-1331
• /
• 1992

본 연구에서는 Fig.3과 같은 다층 퍼셉트론을 사용하기로 한다. 그리고 위 에서 언급한 세가지점에서 다층퍼셉트론을 다시 살펴보아 해결하고자 하는 문제에 맞 도록 다층퍼셉트론을 개선시켜 보기로 한다. 따라서 본 연구의 목적은 제한조건을 갖는 문제를 풀기위한 새로운 형태의 다층퍼셉트론 설계 및 이에 적합한 학습규칙을 적용하여 보다 간단한 구조와 빠른 학습시간을 갖는 신경망을 구성하는데 있다.

• 한국구조물진단유지관리공학회 논문집
• /
• 제6권4호
• /
• pp.169-179
• /
• 2002

This paper presents an efficient algorithm for the estimation of damage location and severity in structure using Probabilistic Neural Network (PNN). Artificial neural network has been being used for damage assessment by many researchers, but there are still some barriers that must be overcome to improve its accuracy and efficiency. The major problems with the conventional neural network are the necessity of many training data for neural network learning and ambiguity in the relation of neural network architecture with convergence of solution. In this paper, PNN is used as a pattern classifier to overcome those problems in the conventional neural network. The basic idea of damage assessment algorithm proposed in this paper is that modal characteristics from a damaged structure are compared with the training patterns which represent the damage in specific element to determine how close it is to training patterns in terms of the probability from PNN. The training pattern that gives a maximum probability implies that the element used in producing the training pattern is considered as a damaged one. The proposed damage assessment algorithm using PNN is applied to a 2-span continuous beam model structure to verify the algorithm.

• 제어로봇시스템학회논문지
• /
• 제5권4호
• /
• pp.419-426
• /
• 1999

A novel motion analysis system is presented in this paper. The proposed system is inspired by processing functions observed in the fly visual system, which detects changes in input light intensities, determines motion on both the local and the wide-field levels. The system has several differences from conventional motion analysis system. First, conventional systems usually focused on matching similar feature or optical flow, but neural network is applied in this system. Back propagation is used by learning method, and Tine Delay Neural Network (TDNN) is also used as analysis method. Second, while conventional systems usually limited on only two frames of sequence, the proposed system accept multiple frames of sequence. The experimental results showed a 94.7% correct rate with a speed of 71.47 milli seconds for real and synthetic images.

• 한국음향학회지
• /
• 제13권4호
• /
• pp.12-19
• /
• 1994

본 논문에서는 음소의 시간 구조 특성만을 다룬 일반적인 TSNN 방법에 음소의 주파수 대역 구조를 포함시킨 신경 회로망을 제안한다. 제안된 신경 회로망에 음소(아, 이, 오, ㅅ, ㅊ, ㅍ, ㄱ, ㅇ, ㄹ, ㅁ)을 학습시켜 인식을 수행한 결과, 시간 인자 특성을 입력으로 음소를 인식한 일반적인 TDNN 방법 과 TSNN 방법보다 본 논문에서 시간과 주파수 인자를 동시에 입력으로 수행한 신경회로망 방법이 약간 더 나은 인식 결과를 보였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.86-86
• /
• 2000

This paper describes a neural network based fuzzy type PID control scheme. The PID controller is being widely used in industrial applications. however, it is difficult to determine the appropriate PID gains for (he nonlinear system control. In this paper, we re-analyzed the fuzzy controller as conventional PID controller structure, and proposed a neural network based fuzzy type PID controller whose scaling factors were adjusted automatically. The value of initial scaling factors of the proposed controller were determined on the basis of the conventional PID controller parameters tuning methods and then they were adjusted by using neural network control techniques. Proposed controller was simple in structure and computational burden was small so that on-line adaptation was easy to apply to. The result of practical experiment on the magnetic levitation system, which is known to be hard nonlinear, showed the proposed controller's excellent performance.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제50권1호
• /
• pp.22-30
• /
• 2001

This paper describes a neural network based fuzzy PID control scheme. The PID controller is being widely used in industrial applications. However, it is difficult to determine the appropriated PID gains in nonlinear systems and systems with long time delay and so on. In this paper, we re-analyzed the fuzzy controller as conventional PID controller structure, and proposed a neural network based self tuning fuzzy PID controller of which output gains were adjusted automatically. The tuning parameters of the proposed controller were determined on the basis of the conventional PID controller parameters tuning methods. Then they were adjusted by using proposed neural network learning algorithm. Proposed controller was simple in structure and computational burden was small so that on-line adaptation was easy to apply to. The experiment on the magnetic levitation system, which is known to be heavily nonlinear, showed the proposed controller's excellent performance.

• Journal of information and communication convergence engineering
• /
• 제2권1호
• /
• pp.46-51
• /
• 2004

The problem of reducing the effect of an unknown disturbance on a dynamical system is one of the most fundamental issues in control design. We propose a robust PID (Proportional Integral Derivative) control method with neural network for improving the performance due to the rejection of an unknown disturbance. The proposed system consists of a model of the plant, a conventional PID controller and a multi-layer neural network, and is composed of two loop; the first loop enables the system to achieve stability of system, the second loop rejects an unknown disturbance. Simulation and experiment results show that the proposed method improves considerably on the performance of the conventional PID control method and the typical IMC method using neural network.

• 대한원격탐사학회지
• /
• 제15권2호
• /
• pp.107-117
• /
• 1999

본 연구에서는 원격탐사를 이용하여 수질 파라미터들을 평가하는데 기존의 다중 회귀나 밴드비 회귀 분석을 이용한 통계적인 방법과 신경망을 이용한 방법을 비교하였다. 사용된 영상은 1996년 3월 18일 대청호 유역의 Landsat TM 영상이며, 30개의 현장 실측치가 위성이 통과하는 시간대에 샘플링되었다. 적용된 신경망은 3개의 층으로 구성된 전향 신경망이며 훈련방법으로는 역전파를 사용하였다. 본 연구에서는 가용한 훈련 데이터 셀이 작으므로 cross-validation 방법이 적용되었다. 비록 기존의 회귀분석에 의한 결과도 어느 정도 유의하게 나왔지만, 신경망에 의한 결과가 훨씬 성공적인 수행을 보여주었다. 신경망을 이용한 수질평가는 신경망이 자료의 비선형적 속성을 잘 반영해주기 때문에 기존의 통계적 기법보다 훨씬 나은 결과를 제공한다고 판단된다.