• International Journal of Control, Automation, and Systems
• /
• 제6권5호
• /
• pp.639-650
• /
• 2008

Electricity price forecasting has become an integral part of power system operation and control. In this paper, a wavelet transform (WT) based neural network (NN) model to forecast price profile in a deregulated electricity market has been presented. The historical price data has been decomposed into wavelet domain constitutive sub series using WT and then combined with the other time domain variables to form the set of input variables for the proposed forecasting model. The behavior of the wavelet domain constitutive series has been studied based on statistical analysis. It has been observed that forecasting accuracy can be improved by the use of WT in a forecasting model. Multi-scale analysis from one to seven levels of decomposition has been performed and the empirical evidence suggests that accuracy improvement is highest at third level of decomposition. Forecasting performance of the proposed model has been compared with (i) a heuristic technique, (ii) a simulation model used by Ontario's Independent Electricity System Operator (IESO), (iii) a Multiple Linear Regression (MLR) model, (iv) NN model, (v) Auto Regressive Integrated Moving Average (ARIMA) model, (vi) Dynamic Regression (DR) model, and (vii) Transfer Function (TF) model. Forecasting results show that the performance of the proposed WT based NN model is satisfactory and it can be used by the participants to respond properly as it predicts price before closing of window for submission of initial bids.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.168.1-168
• /
• 2001

This article describes a novel model-based compensating neural network (NN) model developed to be used in our active binocular head controller, which addresses both the kinematics and dynamics aspects in trying to precisely track a moving object of interest to keep it in view. The compensating NN model is constructed using two classes of self-tuning neural models: namely Neural Gas (NG) algorithm and SoftMax function networks. The resultant servo controller is shown to be able to handle the tracking problem with a minimum knowledge of the dynamic aspects of the system.

• Journal of information and communication convergence engineering
• /
• 제6권4호
• /
• pp.448-453
• /
• 2008

This study describes design and development techniques of estimation models for process modeling. One case study is undertaken to design a model using standard gas furnace data. Neural networks (NN) and genetic programming (GP) are each employed to model the crucial relationships between input factors and output responses. In the case study, two models were generated by using 70% training data and evaluated by using 30% testing data for genetic programming and neural network modeling. The model performance was compared by using RMSE values, which were calculated based on the model outputs. The average RMSE for training and testing were 0.8925 (training) and 0.9951 (testing) for the NN model, and 0.707227 (training) and 0.673150 (testing) for the GP model, respectively. As concern the results, the NN model has a strong advantage in model training (using the all data for training), and the GP model appears to have an advantage in model testing (using the separated data for training and testing). The performance reproducibility of the GP model is good, so this approach appears suitable for modeling physical fabrication processes.

• Smart Structures and Systems
• /
• 제29권5호
• /
• pp.693-703
• /
• 2022

This study investigates an effective approach to stabilize nonlinear systems. To ensure the asymptotic nonlinear stability in nonlinear discrete-time systems, the present study presents controller for an EBA (Evolved Bat Algorithm) NN (fuzzy neural network) in the algorithm. In fuzzy evolved NN modeling, the auxiliary circuit with high frequency LDI (linear differential inclusions) and NN model representation is developed for the nonlinear arbitrary dynamics. An example is utilized to demonstrate the system more robust compared with traditional control systems.

• Composites Research
• /
• 제33권3호
• /
• pp.108-114
• /
• 2020

Piecewise Integrated Composite (PIC) 보는 하중 유형에 따라 구간을 나누어, 각 구간마다 하중 유형에 강한 복합재료의 적층 순서를 배열한 보이다. 본 연구는 PIC 보의 구간을 머신 러닝의 일종인 k-NN(k-Nearest Neighbor) 분류를 통해 나누어 기존에 제시되었던 PIC 보에 비해 우수한 굽힘 특성을 갖게 하는 것이 목적이다. 먼저, 알루미늄 보의 3점 굽힘 해석을 통하여 참조점에서의 3축 특성(Triaxiality) 값 데이터를 얻었고, 이를 통해 인장, 전단, 압축의 레이블을 가진 학습 데이터가 만들어진다. 학습 데이터를 통해 각 면마다 독립적인 k-NN 분류 모델을 구성하는 방법(Each plane)과 전체 면에 대한 k-NN 분류 모델을 구성하는 방법(one part)을 이용하여 k-NN 분류 모델을 생성하였고, 하이퍼파라미터의 튜닝을 통하여 다양한 하중 충실도를 도출하였다. 가장 높은 하중 충실도를 가진 k-NN 분류 모델을 기반으로 보를 매핑(mapping)하였고, PIC 보에 대하여 유한요소 해석을 진행한 결과, 기존에 제시되었던 PIC 보에 비해 최대하중과 흡수 에너지가 커지는 특성을 보였다. 하중 충실도를 수동으로 조절하여 100%로 만든 PIC 보와 비교하였을 때, 최대하중과 흡수에너지가 미소한 차이가 나타났으며 이는 타당한 하중 충실도로 보여진다.

• Journal of Power Electronics
• /
• 제6권4호
• /
• pp.340-346
• /
• 2006

A high efficiency drive technique for synchronous reluctance motors (SynRM) using a neural network (NN) is presented in this paper. High efficiency drive condition depends on the mathematical model of SynRM. A NN is employed as an adaptive model of SynRM. The proposed high efficiency drive technique does not require an accurate mathematical model of SynRM. Moreover, the proposed method shows robustness against machine parameter variations because the training algorithm of the NN is executed on-line. The usefulness of the proposed method is confirmed through experimentation.

• Structural Engineering and Mechanics
• /
• 제23권2호
• /
• pp.177-193
• /
• 2006

Determining the hysteretic energy demand and dissipation capacity and level of damage of the structure to a predefined earthquake ground motion is a highly non-linear problem and is one of the questions involved in predicting the structure's response for low-performance levels (life safe, near collapse, collapse) in performance-based earthquake resistant design. Neural Network (NN) analysis offers an alternative approach for investigation of non-linear relationships in engineering problems. The results of NN yield a more realistic and accurate prediction. A NN model can help the engineer to predict the seismic performance of the structure and to design the structural elements, even when there is not adequate information at the early stages of the design process. The principal aim of this study is to develop and test multi-layered feedforward NNs trained with the back-propagation algorithm to model the non-linear relationship between the structural and ground motion parameters and the hysteretic energy demand in steel moment resisting frames. The approach adapted in this study was shown to be capable of providing accurate estimates of hysteretic energy demand by using the six design parameters.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 1998년도 가을 학술발표회 논문집
• /
• pp.439-446
• /
• 1998

One of the most important roles in the nonlinear dynamic structural analysis is to select a proper ground excitation, which dominates the response of a structure. Because of the lack of recorded accelerograms in Korea, a stochastic model of ground excitation with various dynamic properties rather than recorded accelerograms is necessarily required. If all information is not available at site, the information from other sites with similar features can be used by the procedure of seismic hazard analysis. Eliopoulos and Wen identified the parameters of the ground motion model by the empirical relations or expressions developed by Trifunac and Lee. Because the relations used in the parameter identification are largely empirical, it is required to apply the artificial neural networks instead of the empirical model. Additionally, neural networks have the advantage of the empirical model that it can continuously re-train the new recorded data, so that it can adapt to the change of the enormous data. Based on the redefined traditional processes, three neural-networks-based models (FAS_NN, PSD_NN and INT_NN) are proposed to individually substitute the Fourier amplitude spectrum, the parameter identification of power spectral density function and intensity function. The paper describes the first half of the research for the development of Neural-Networks-based model for the generation of an Artificial earthquake and a Response Spectrum(NNARS).

• 한국항공우주학회지
• /
• 제35권11호
• /
• pp.990-998
• /
• 2007

불확실한 비선형 다중입출력 시스템에 대해서 신경회로망을 이용한 적응출력피드백제어기법이 제안되었다. 역변환 기반의 제어입력으로부터 불확실한 비선형성을 분리하기 위해 변형된 운동 역변환 모델(Modified Dynamic Inversion Model, MDIM)이 도입되었다. MDIM은 근사된 운동 역변환 모델과 역변환 모델 오차로 구성되었고 한 개의 신경회로망이 MDIM을 보상하는데 적용되었다. 여기서 신경회로망의 출력은 필터링된 근사오차 기반의 제어기를 증대시킨다. 추적성능과 종국적 유계성(ultimate boundedness)을 보장하기 위해 리야프노프의 직접방법(Lyapunov's direct method)으로부터 유도된 온라인 가중치 적응법칙이 이용되었다. 수치적 시뮬레이션을 통해 본 논문의 타당성을 검증하였다.

• Computers and Concrete
• /
• 제28권6호
• /
• pp.635-642
• /
• 2021

In this paper, soft learning techniques are used to predict the ultimate torsional capacity of Reinforced Concrete beams strengthened with Fiber Reinforced Polymer. Soft computing techniques, namely Artificial Neural Network, trained by various back propagation algorithms, and Particle Swarm Optimization (PSO) algorithm, have been used to model and predict the torsional strength of Reinforced Concrete beams strengthened with Fiber Reinforced Polymer. The performance of each model has been evaluated by using statistical parameters such as coefficient of determination (R²), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The hybrid PSO NN model resulted in an R² of 0.9292 with an RMSE of 5.35 for training and an R² of 0.9328 with an RMSE of 4.57 for testing. Another model, ANN BP, produced an R² of 0.9125 with an RMSE of 6.17 for training and an R² of 0.8951 with an RMSE of 5.79 for testing. The results of the PSO NN model were in close agreement with the experimental values. Thus, the PSO NN model can be used to predict the ultimate torsional capacity of RC beams strengthened with FRP with greater acceptable accuracy.