• 한국경영과학회지
• /
• 제28권2호
• /
• pp.105-127
• /
• 2003

In this paper, four different data mining techniques, two neural networks and two statistical modeling techniques, are compared in terms of prediction accuracy in the context of bankruptcy prediction. In business setting, how to accurately detect the condition of a firm has been an important event in the literature. In neural networks, Backpropagation (BP) network and the Kohonen self-organizing feature map, are selected and compared each other while in statistical modeling techniques, discriminant analysis and logistic regression are also performed to provide performance benchmarks for the neural network experiment. The findings suggest that the BP network is a better choice among the data mining tools compared. This paper also identified some distinctive characteristics of Kohonen self-organizing feature map.

• Nuclear Engineering and Technology
• /
• 제55권9호
• /
• pp.3409-3416
• /
• 2023

Licensing the next-generation of nuclear reactor designs requires extensive use of Modeling and Simulation (M&S) to investigate system response to many operational conditions, identify possible accidental scenarios and predict their evolution to undesirable consequences that are to be prevented or mitigated via the deployment of adequate safety barriers. Deep Learning (DL) and Artificial Intelligence (AI) can support M&S computationally by providing surrogates of the complex multi-physics high-fidelity models used for design. However, DL and AI are, generally, low-fidelity 'black-box' models that do not assure any structure based on physical laws and constraints, and may, thus, lack interpretability and accuracy of the results. This poses limitations on their credibility and doubts about their adoption for the safety assessment and licensing of novel reactor designs. In this regard, Physics Informed Neural Networks (PINNs) are receiving growing attention for their ability to integrate fundamental physics laws and domain knowledge in the neural networks, thus assuring credible generalization capabilities and credible predictions. This paper presents the use of PINNs as surrogate models for accidental scenarios simulation in Nuclear Power Plants (NPPs). A case study of a Loss of Heat Sink (LOHS) accidental scenario in a Nuclear Battery (NB), a unique class of transportable, plug-and-play microreactors, is considered. A PINN is developed and compared with a Deep Neural Network (DNN). The results show the advantages of PINNs in providing accurate solutions, avoiding overfitting, underfitting and intrinsically ensuring physics-consistent results.

• 지구물리와물리탐사
• /
• 제26권3호
• /
• pp.114-125
• /
• 2023

편미분방정식의 해를 구하기 위한 여러 수치해법들의 한계와 순수 데이터 기반 기계학습의 단점을 극복하기 위해 물리정보신경망(physics-informed neural network, PINN)이 제안되었다. 물리정보신경망은 편미분방정식을 손실함수 구성에 직접 활용하여 기계학습 훈련에 물리적 제약을 주는 기법으로 파동방정식 모델링에도 활용될 수 있다. 그러나 물리정보신경망을 이용하여 파동방정식을 풀기 위해서는 신경망 훈련 시 입력에 대한 2차 미분이 수행되어야 하고, 그 결과로 출력되는 파동장은 복잡한 역학적 현상들을 포함하고 있어 섬세한 전략이 필요하다. 이 해설 논문에서는 물리정보신경망의 기본 개념을 설명하고 파동방정식 모델링에 활용하기 위한 고려사항들에 대해 고찰하였다. 이러한 고려사항에는 공간좌표 정규화, 활성함수 선정, 물리손실 추가 전략이 포함된다. 훈련자료의 공간좌표를 정규화한 후 사용하면 파동방정식 모델링을 위한 신경망 훈련에서 초기 조건이 더 정확하게 반영되는 것을 수치 실험을 통해 보였다. 또한 신경망을 통한 파동장 예측에 가장 적절한 활성함수를 선정하기 위해 여러 함수들의 특성을 비교했다. 특성 비교는 각 활성함수들의 입력자료에 대한 미분과 수렴성을 중심으로 이루어졌다. 마지막으로 신경망 훈련 중 손실함수에 물리손실을 추가하는 두가지 시나리오의 결과를 비교하였다. 수치 실험을 통해 훈련 초기부터 물리손실을 활용하는 전략보다 초기 훈련단계 이후부터 물리손실을 적용하는 커리큘럼 기반 학습전략이 효과적이라는 결과를 도출했다. 추가로 이 결과를 물리손실을 전혀 사용하지 않은 훈련 결과와 비교하여 PINN기법의 효과를 확인하였다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 정보 및 제어부문
• /
• pp.155-157
• /
• 2005

We propose a new category of hybrid multi-layer neural networks with hetero nodes such as Fuzzy Set based Polynomial Neurons (FSPNs) and Polynomial Neurons (PNs). These networks are based on a genetically optimized multi-layer perceptron. We develop a comprehensive design methodology involving mechanisms of genetic optimization and genetic algorithms, in particular. The augmented genetically optimized HFPNN (namely gHFPNN) results in a structurally optimized structure and comes with a higher level of flexibility in comparison to the one we encounter in the conventional HFPNN. The GA-based design procedure being applied at each layer of HFPNN leads to the selection of preferred nodes (FPNs or PNs) available within the HFPNN. In the sequel, two general optimization mechanisms are explored. First, the structural optimization is realized via GAs whereas the ensuing detailed parametric optimization is carried out in the setting of a standard least square method-based learning. The performance of the gHFPNNs quantified through experimentation where we use a number of modeling benchmarks-synthetic and experimental data already experimented with in fuzzy or neurofuzzy modeling.

• Transactions on Control, Automation and Systems Engineering
• /
• 제2권3호
• /
• pp.189-195
• /
• 2000

The mathematical solutions of the stability convergence are important problems in system control. In this paper such problems are analyzed and resolved for system control using multilayer neural networks. We describe an algorithm to control an unknown nonlinear system with a disturbance, using a multilayer neural network. We include a disturbance among the modeling error, and the weight update rules of multilayer neural network are derived to satisfy Lyapunov stability. The overall control system is based upon the feedback linearization method. The weights of the neural network used to approximate a nonlinear function are updated by rules derived in this paper . The proposed control algorithm is verified through computer simulation. That is as the weights of neural network are updated at every sampling time, we show that the output error become finite within a relatively short time.

• 제어로봇시스템학회논문지
• /
• 제13권5호
• /
• pp.422-428
• /
• 2007

This paper presents a neural network based adaptive control approach to a reconfigurable flight control law that keeps handling qualities in the presence of faults or failures to the control surfaces of an aircraft. This approach removes the need for system identification for control reallocation after a failure and the need for an accurate aerodynamic database for flight control design, thereby reducing the cost and time required to develope a reconfigurable flight controller. Neural networks address the problem caused by uncertainties in modeling an aircraft and pseudo control hedging deals with the nonlinearity in actuators and the reconfiguration of a flight controller. The effect of the reconfigurable flight control law is illustrated in results of a nonlinear simulation of an unmanned aerial vehicle Durumi-II.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 학술대회 논문집 정보 및 제어부문
• /
• pp.322-324
• /
• 2005

In this paper, we introduce the advanced Self-Organizing Fuzzy Polynomial Neural Network based on optimized FPN by evolutionary algorithm and discuss its comprehensive design methodology involving mechanisms of genetic optimization, especially genetic algorithms (GAs). The proposed model gives rise to a structurally and parametrically optimized network through an optimal parameters design available within Fuzzy Polynomial Neuron(FPN) by means of GA. Through the consecutive process of such structural and parametric optimization, an optimized and flexible the proposed model is generated in a dynamic fashion. The performance of the proposed model is quantified through experimentation that exploits standard data already used in fuzzy modeling. These results reveal superiority of the proposed networks over the existing fuzzy and neural models.

• 제어로봇시스템학회논문지
• /
• 제14권12호
• /
• pp.1226-1231
• /
• 2008

A design of attitude controller for a tiltrotor is presented augmenting L1 adaptive control, neural networks, and feedback linearization. The neural networks compensate for the modeling error caused by the lack of knowledge of tiltrotor dynamics while the L1 adaptive control allows high adaptation gains in adaptation laws thereby, satisfying tracking performance requirement. The efficacy of this control methodology is illustrated in high-fidelity nonlinear simulation of a tiltrotor by flying the tiltrotor in different flight modes from where the L1 adaptive controller with neural networks is originally designed for.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1993년도 하계학술대회 논문집 A
• /
• pp.224-226
• /
• 1993

Multi-layered neural networks offer an exciting alternative for modelling complex non-liner systems. This paper investigates the identification of continuous time nonliner system using neural networks with a single hidden layer. The digital low - pass filter are introduced to avoid direct approximation of system derivatives from sampled data. Using a pre-designed digital low pass filter, an approximated discrete-time estimation model is constructed easily. A continuous approximation liner model is first estimated from sampled input-out signals. Then the modeling error due to the nonlinearity is decreased by a compensator using neural network. Simulation results are given to demonstrate the effective of the proposed method.

• 한국자동차공학회논문집
• /
• 제7권6호
• /
• pp.258-269
• /
• 1999

It is important to evaluate fatigue damage of in-service material in respect to assure safety and remaining fatigue life in structure and mechanical components under cyclic load . Fatigue damage is represented by mathematical modelling with crack growth rate da/dN and cycle ration N/Nf and is detected by X-ray diffraction and ultrasonic wave method etc. But this is estimated generally by single parameter but influenced by many test conditions The characteristics of it indicates fatigue damage has complex fracture mechanism. Therefore, in this study we propose that back-propagation neural networks on the basis of ration of X-ray half-value breath B/Bo, fractal dimension Df and fracture mechanical parameters can construct artificial intelligent networks estimating crack growth rate da/dN and cycle ratio N/Nf without regard to stress amplitude Δ $\sigma$.