• 대한전기학회논문지:시스템및제어부문D
• /
• 제53권1호
• /
• pp.40-49
• /
• 2004

In this paper, we propose a new architecture of Genetic Algorithms(GAs)-based Self-Organizing Polynomial Neural Networks(SOPNN), discuss a comprehensive design methodology and carry out a series of numeric experiments. The conventional SOPNN is based on the extended Group Method of Data Handling(GMDH) method and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons (or nodes) located in each layer through a growth process of the network. Moreover it does not guarantee that the SOPNN generated through learning has the optimal network architecture. But the proposed GA-based SOPNN enable the architecture to be a structurally more optimized network, and to be much more flexible and preferable neural network than the conventional SOPNN. In order to generate the structurally optimized SOPNN, GA-based design procedure at each stage (layer) of SOPNN leads to the selection of preferred nodes (or PNs) with optimal parameters- such as the number of input variables, input variables, and the order of the polynomial-available within SOPNN. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between approximation and generalization (predictive) abilities of the model. A detailed design procedure is discussed in detail. To evaluate the performance of the GA-based SOPNN, the model is experimented with using two time series data (gas furnace and NOx emission process data of gas turbine power plant). A comparative analysis shows that the proposed GA-based SOPNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 추계학술대회 논문집 학회본부 D
• /
• pp.806-808
• /
• 2000

We propose a new Dynamic Multilayer Perceptorn(DMP) architecture for optimal model identification of complex and nonlinear system in this paper. The proposed DMP scheme is presented as the generic and advanced type based on the GMDH(Group Method of Data Handling) method for the limitation of GMDH under only two system input variables. It is worth stressing that the number of the layers and the nodes in each layer of the DMP are not predetermined, unlike in the case of the popular multilayer perceptron structure, but these are generated in a dynamic manner. The experimental part of the study comes with representative nonlinear static system. Comparative analysis is included and shows that a new DMP can produce the model with higher accuracy than previous other works.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제49권7호
• /
• pp.378-388
• /
• 2000

In this paper, we propose Fuzzy Polynomial Neural Networks(FPNN) based on Polynomial Neural Networks(PNN) and Fuzzy Neural Networks(FNN) for model identification of complex and nonlinear systems. The proposed FPNN is generated from the mutually combined structure of both FNN and PNN. The one and the other are considered as the premise part and consequence part of FPNN structure respectively. As the consequence part of FPNN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. FPNN is available effectively for multi-input variables and high-order polynomial according to the combination of FNN with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. As the premise part of FPNN, FNN uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. And we use two kinds of FNN structure according to the division method of fuzzy space of input variables. One is basic FNN structure and uses fuzzy input space divided by each separated input variable, the other is modified FNN structure and uses fuzzy input space divided by mutually combined input variables. In order to evaluate the performance of proposed models, we use the nonlinear function and traffic route choice process. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously. And also performance index related to the approximation and prediction capabilities of model is evaluated and discussed.

• International Journal of Control, Automation, and Systems
• /
• 제2권4호
• /
• pp.423-434
• /
• 2004

In this paper, we introduce a new topology of Self-Organizing Polynomial Neural Networks (SOPNN) based on genetically optimized Multi-Layer Perceptron (MLP) and discuss its comprehensive design methodology involving mechanisms of genetic optimization. Let us recall that the design of the 'conventional' SOPNN uses the extended Group Method of Data Handling (GMDH) technique to exploit polynomials as well as to consider a fixed number of input nodes at polynomial neurons (or nodes) located in each layer. However, this design process does not guarantee that the conventional SOPNN generated through learning results in optimal network architecture. The design procedure applied in the construction of each layer of the SOPNN deals with its structural optimization involving the selection of preferred nodes (or PNs) with specific local characteristics (such as the number of input variables, the order of the polynomials, and input variables) and addresses specific aspects of parametric optimization. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between the approximation and generalization (predictive) abilities of the model. To evaluate the performance of the GA-based SOPNN, the model is experimented using pH neutralization process data as well as sewage treatment process data. A comparative analysis indicates that the proposed SOPNN is the model having higher accuracy as well as more superb predictive capability than other intelligent models presented previously.reviously.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2002년도 추계학술대회 및 정기총회
• /
• pp.195-198
• /
• 2002

고농도오존이 발생되는 원인과 환경적 요인의 상호관계를 모델링하기 위해 신경회로 망과 같은 지능제어 기법들이 많이 적용되어 왔다 분석과 모델링을 위해 유전자 알고리즘과 같은 최적화 방법을 적용하기도 하지만, 고농도 오존이 발생되는 메커니즘이 매우 복잡하고, 비선형적이며, 패턴파악이 어렵기 때문에 고농도 오존의 예측 모델링에는 여전히 문제점이 있다 따라서 본 논문에서는 신뢰수준과 신뢰구간을 이용하여 초농도 오존을 예측할 수 있는 모델링 방법을 서술하였다 예측값의 신뢰수준의 평가는 예측에 대한 실측값을 구하여 신뢰구간내의 데이터의 개수를 파악함으로써 신뢰성을 평가할 수 있다. 또한 이 테스트는 우리가 가지고 있지 않은 데이터에 대한 유효성을 평가하는데 적용될 수 있다 그리고 본 논문에서는 GMDH(Group Method of data handling)의 전형적인 알고리즘에 바탕을 두고 있는 DPNN(Dynamic Polynomial Neural Network)를 이용하여 예측 모델을 구성하였다. DPNN은 데이터 해석이 용이하고 비선형적인 동적 시스템 예측에 유용하게 적용될 수 있는 장점을 가지고 있다.

• Geomechanics and Engineering
• /
• 제33권3호
• /
• pp.231-242
• /
• 2023

This paper represents a comparative study in which Gene Expression Programming (GEP), Group Method of Data Handling (GMDH), and multiple linear regressions (MLR) were utilized to derive new equations for the prediction of time-dependent bearing capacity of pile foundations driven in cohesive soil, technically called pile set-up. This term means that many piles which are installed in cohesive soil experience a noticeable increase in bearing capacity after a specific time. Results of researches indicate that side resistance encounters more increase than toe resistance. The main reason leading to pile setup in saturated soil has been found to be the dissipation of excess pore water pressure generated in the process of pile installation, while in unsaturated conditions aging is the major justification. In this study, a comprehensive dataset containing information about 169 test piles was obtained from literature reviews used to develop the models. to prepare the data for further developments using intelligent algorithms, Data mining techniques were performed as a fundamental stage of the study. To verify the models, the data were randomly divided into training and testing datasets. The most striking difference between this study and the previous researches is that the dataset used in this study includes different piles driven in soil with varied geotechnical characterization; therefore, the proposed equations are more generalizable. According to the evaluation criteria, GEP was found to be the most effective method to predict set-up among the other approaches developed earlier for the pertinent research.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1999년도 하계학술대회 논문집 G
• /
• pp.2856-2858
• /
• 1999

In this paper, the FPNN(Fuzzy Polynomial Neural Networks) algorithm with multi-layer fuzzy inference structure is proposed for the model identification of a complex nonlinear system. The FPNN structure is generated from the mutual combination of PNN (Polynomial Neural Network) structure and fuzzy inference method. The PNN extended from the GMDH(Group Method of Data Handling) uses several types of polynomials such as linear, quadratic and modifled quadratic besides the biquadratic polynomial used in the GMDH. In the fuzzy inference method, simplified and regression polynomial inference method which is based on the consequence of fuzzy rule expressed with a polynomial such as linear, quadratic and modified quadratic equation are used Each node of the FPNN is defined as a fuzzy rule and its structure is a kind of fuzzy-neural networks. Gas furnace data used to evaluate the performance of our proposed model.

• Geomechanics and Engineering
• /
• 제14권3호
• /
• pp.257-269
• /
• 2018

Analysis of slope stability failures, as one of the complex natural hazards, is one of the important research issues in the field of civil engineering. Present paper adopts and investigates four soft computing-based techniques for this problem: Patient Rule-Induction Method (PRIM), M5' algorithm, Group Method of data Handling (GMDH) and Multivariate Adaptive Regression Splines (MARS). A comprehensive database consisting of 168 case histories is used to calibrate and test the developed models. Six predictive variables including slope height, slope angle, bulk density, cohesion, angle of internal friction, and pore water pressure ratio were considered to generate new models. The results of test studies are used for feasibility, effectiveness and practicality comparison of techniques with each other, and with the other available well-known methods in the literature. Results show that all methods not only are feasible but also result in better performance than previously developed soft computing based predictive models and tools. It is shown that M5' and PRIM algorithms are the most effective and practical prediction models.

• 한국지능시스템학회논문지
• /
• 제14권5호
• /
• pp.660-665
• /
• 2004

In this study, a new architecture and comprehensive design methodology of genetically optimized Multi-layer Fuzzy Neural Networks (gMFNN) are introduced and a series of numeric experiments are carried out. The gMFNN architecture results from a synergistic usage of the hybrid system generated by combining Fuzzy Neural Networks (FNN) with Polynomial Neural Networks (PNN). FNN contributes to the formation of the premise part of the overall network structure of the gMFNN. The consequence part of the gMFNN is designed using PNN. The optimization of the FNN is realized with the aid of a standard back-propagation learning algorithm and genetic optimization. The development of the PNN dwells on the extended Group Method of Data Handling (GMDH) method and Genetic Algorithms (GAs). To evaluate the performance of the gMFNN, the models are experimented with the use of a numerical example.

• 한국지능시스템학회논문지
• /
• 제15권7호
• /
• pp.846-851
• /
• 2005

본 논문은 최적 탐색 알고리즘인 유전자 알고리즘을 이용하여 다항식 뉴럴네트워크(Polynomial Neural Networks : PNN)의 최적 설계가 그 목적이다. 기존의 다항식 뉴럴네트워크는 확장된 GMDH(Group Method of Data Handling) 방법에 기반을 두며, 네트워크의 성장과정을 통하여 각 층의 다항식뉴런(혹은 노드)에서 고정된 (설계자에 의해 미리 선택된) 노드 입력들의 수뿐만 아니라 다항식 차수(1차, 2차, 그리고 수정된 2차식)를 이용하였다. 더구나, 그 방법은 학습을 통해 생성된 PNN이 최적 네트워크 구조를 가진다는 것을 보증하지 못한다. 그러나, 제안된 GA-based PNN 모델은 다음의 파라미터들- 즉 입력변수의 수, 입력변수, 및 다항식 차수-을 유전자 알고리즘을 이용하여 선택 동조함으로써 그 구조를 구조적으로 더 최적화된 네트워크가 되도록 하고, 기존의 PNN보다 훨씬 더 유연하고, 선호된 뉴럴 네트워크가 되도록 한다. 하중계수를 가진 합성성능지수가 그 모델의 근사화 및 일반화(예측) 능력 사이의 상호 균형을 얻기 위해 제안된다. GA-based PNN의 성능을 평가하기 위해 그 모델은 가스 터빈 발전소의 NOx 배출 공정 데이터로 실험된다. 비교해석은 제안된 GA-based PNN이 앞서 나타난 다른 지능모델보다 더 우수한 예측능력뿐만 아니라 높은 정확성을 가진 모델임을 보인다.