• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.2879-2881
• /
• 2005

본 논문에서는 뉴로-퍼지 모델의 전제부 소속함수의 새로운 학습방법을 통한 모델링 기법을 제안한다. 모델의 크기와 학습시간을 줄이는 기법으로 클러스터링 기법을 이용한 모델의 초기 파라미터 결정 방법이 있다. 이는 클러스터링 후 이들 파라미터를 다시 모델에 적용하여 모델을 학습하는 순차적 방법으로써 모델의 학습이 끝난 후의 전제부 파라미터가 클러스터링 파라미터와 연관성을 가지지 못하는 경우가 발생하였다. 또한 오차미분 기반 학습에서는 전제부 초기치가 국부적 최적해에서 벋어나지 못하는 문제점을 가지고 있다. 본 논문에서는 자율적으로 클러스터의 수를 추정하며 이들 파라미터를 최적화하며 이를 이용하여 뉴로-퍼지 모델의 학습을 실시하는 학습기법을 제안하였다. 제안된 방법에서는 기존의 오차미분 기반 학습을 클러스터링 기반 학습으로 확장하였으며 이를 이용한 모델의 성능을 기존의 연구결과와 비교하여 우수성을 보인다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.8
• /
• pp.695-700
• /
• 2001

We design the intelligent diagnosis system for deciding on operation of TMS Analysis in this paper. We propose the method to model the neuro-fuzzy model for diagnosing the operation state of analyzer by using input and output signals of TMS and Expert's experiment data. Validity of the proposed system is asserted by numerical simulation.

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.182-183
• /
• 2007

본 논문에서는 전형적인 비선형 회귀문제를 다루기 위해 슈뢰딩거 방정식에 의해 표현되는 Hilbert공간에서 수행되는 Quantum 클러스터링과 Mountain 함수를 이용하여, 수치적인 입출력데이터로부터 TSK 형태의 자동적인 퍼지 if-then 규칙의 생성방법을 제안한다. 여기서 슈뢰딩거 방정식은 분석적으로 확률함수로부터 유도되어질 수 있는 포텐셜 함수를 포함한다. 이 포텐셜의 최소점들은 데이터의 특성을 포함하는 클러스터 중심들과 관련되어진다. 그러나 이들 클러스터 중심들은 데이터의 수와 같으므로 퍼지 규칙을 생성하기 어려울 뿐만 아니라 수렴속도가 느린 문제점을 가지고 있다. 이러한 문제점들을 해결하기 위해서, 본 논문에서는 밀도 척도에 기초한 클러스터 중심의 근사적인 추정에 대해 간단하면서 효과적인 Mountain 함수를 이용하여 효과적인 클러스터 중심을 얻음과 동시에 적응 뉴로-퍼지 네트워크의 자동적인 퍼지 규칙을 생성하도록 한다. 자동차 MPG 예측문제에 대한 시뮬레이션 결과는 제안된 방법이 기존 문헌에서 제시한 예측성능보다 더 좋은 특성을 보임을 알 수 있었다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.960-963
• /
• 1993

We focus our attention on grading of table meat in accordance with the standard of Japan Meat Grading Association, and construct a beef grading system by image processing. For image processing of beef grading, it needs some techniques such as a shading correction, separation of color image data, and classification of color image data into some grades, for the system construction. However, there are various kinds of weak points in usually used methods for these techniques. Then the authors propose and introduce new approaches using Neural networks and fuzzy inference for the techniques above mentioned, which is very convenient and ensure the high precision.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• v.1
• /
• pp.115-120
• /
• 2006

This article propose a new idea of developing a hybrid scheme to achieve faster INS alignment with higher accuracy using a novel procedure to estimate the initial attitude angles that combines a Kalman filter and Adaptive Neuro-Fuzzy Inference System architecture. A tactical grade inertial measurement unit was applied to verify the performance of proposed scheme in this study. The preliminary results indicated the outstanding improvements in both time consumption for fine alignment process and accuracy of estimated attitude angles, especially in heading angles. In general, the improvement in terms of time consumption and the accuracy of estimated attitude estimated accuracy reached 80% and 70% respectively during alignment process after compensating the attitude angles estimated by an extended Kalman filter with 15 states using proposed approach. It is worth mentioned that the proposed approach can be implemented in general real time navigation applications.

• Journal of information and communication convergence engineering
• /
• v.2 no.1
• /
• pp.36-39
• /
• 2004

In this paper, a method of improving the learning speed and convergence rate is proposed to exploit the advantages of artificial neural networks and neuro-fuzzy systems. This method is applied to the XOR problem, n bit parity problem, which is used as the benchmark in the field of pattern recognition. The method is also applied to the recognition of digital image for practical image application. As a result of experiment, it does not always guarantee convergence. However, the network showed considerable improvement in learning time and has a high convergence rate. The proposed network can be extended to any number of layers. When we consider only the case of the single layer, the networks had the capability of high speed during the learning process and rapid processing on huge images.

• International Journal of Reliability and Applications
• /
• v.16 no.1
• /
• pp.35-53
• /
• 2015

This investigation deals with reliability and sensitivity analysis of a repairable embedded system with standby wherein repairman takes multiple vacations. The hardware system consists of 'M' operating and 'S' standby components. The repairman can leave for multiple vacations of random length during its idle time. Whenever any operating unit fails, it is immediately replaced by a standby unit if available. Moreover, governing equations of an embedded system are constructed using appropriate birth-death rates. The vacation and repair time of repairman are exponentially distributed. The matrix method is used to find the steady-state probabilities of the number of failed components in the embedded system as well as other performance measures. Reliability indexes are presented. Further, numerical experiments are carried out for various system characteristics to examine the effects of different parameter. Using a special class of neuro-fuzzy systems i.e. Adaptive Network-based Fuzzy Interference Systems (ANFIS), we also approximate various performance measures. Finally, the conclusions and future research directions are provided.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.6
• /
• pp.949-963
• /
• 2017

This study present an efficient way of operating drainage pump station as part of nonstructural measures for reducing urban flood damage. The water level in the drainage pump station was forecast using Neuro-Fuzzy and then operation rule of the drainage pump station was determined applying the genetic algorithm method based on the predicted inner water level. In order to reflect the topographical characteristics of the drainage area when constructing the Neuro-Fuzzy model, the model considering spatial parameters was developed. Also, the model was applied a penalty type of genetic algorithm so as to prevent repeated stops and operations while lowering my highest water level. The applicability of the development model for the five drainage pump stations in the Mapo drainage area was verified. It is considered to be able to effectively manage urban drainage facilities in the development of these operating rules.

• 한국정보통신설비학회:학술대회논문집
• /
• 2007.08a
• /
• pp.181-184
• /
• 2007

Up to now, a lot of houses, roads and other urban facilities have been damaged by natural disasters such as flash floods and landslides. It is reported that the size and frequency of disasters are growing greatly due to global warming. In order to mitigate such disaster, flood forecasting and alerting systems have been developed for the Han river, Geum river, Nak-dong river and Young-san river. These systems, however, do not help small municipal departments cope with the threat of flood. In this study, a real-time urban flood forecasting service (U-FFS) is developed for ubiquitous computing city which includes small river basins. A test bed is deployed at Tan-cheon in Gyeonggido to verify U-FFS. Wireless sensors such as rainfall gauge and water lever gauge are installed to develop hydrologic forecasting model and CCTV camera systems are also incorporated to capture high definition images of river basins. U-FFS is based on the ANFIS (Adaptive Neuro-Fuzzy Inference System) that is data-driven model and is characterized by its accuracy and adaptability. It is found that U-FFS can forecast the water level of outlet of river basin and provide real-time data through internet during heavy rain. It is revealed that U-FFS can predict the water level of 30 minutes and 1 hour later very accurately. Unlike other hydrologic forecasting model, this newly developed U-FFS has advantages such as its applicability and feasibility. Furthermore, it is expected that U-FFS presented in this study can be applied to ubiquitous computing city (U-City) and/or other cities which have suffered from flood damage for a long time.

• Journal of Korea Water Resources Association
• /
• v.38 no.7 s.156
• /
• pp.575-584
• /
• 2005

The judgement model to warn of possible pollution accident is constructed by multi-perceptron, multi layer neural network, neuro-fuzzy and it is trained stability, notice, and warming situation due to developed standard axis. The water quality forecasting model is linked to the runoff forecasting model, and joined with the judgement model to warn of possible pollution accident, which completes the artificial intelligence warning system. And GUI (Graphic User Interface) has been designed for that system. GUI screens, in order of process, are main page, data edit, discharge forecasting, water quality forecasting, warming system. The application capability of the system was estimated by the pollution accident scenario. Estimation results verify that the artificial intelligence warning system can be a reasonable judgement of the noized water pollution data.