• Wind and Structures
• /
• v.38 no.2
• /
• pp.147-160
• /
• 2024

The aerodynamic force is a significant component that influences the stability and safety of structures. It has unstable properties and depends on computer precision, making its long-term prediction challenging. Accurately estimating the aerodynamic traits of structures is critical for structural design and vibration control. This paper establishes an unsteady aerodynamic time series prediction model using Long Short-Term Memory (LSTM) network. The unsteady aerodynamic force under varied Reynolds number and angles of attack is predicted by the LSTM model. The input of the model is the aerodynamic coefficients of the 1 to n sample points and output is the aerodynamic coefficients of the n+1 sample point. The model is predicted by interpolation and extrapolation utilizing Unsteady Reynolds-average Navier-Stokes (URANS) simulation data of flow around a circular cylinder, square cylinder and airfoil. The results illustrate that the trajectories of the LSTM prediction results and URANS outcomes are largely consistent with time. The mean relative error between the forecast results and the original results is less than 6%. Therefore, our technique has a prospective application in unsteady aerodynamic force prediction of structures and can give technical assistance for engineering applications.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.12
• /
• pp.1772-1781
• /
• 2017

In this study, the design methodology for alleviating the overfitting problem of Polynomial Neural Networks(PNN) is realized with the aid of two kinds techniques such as L2 regularization and Sum of Squared Coefficients (SSC). The PNN is widely used as a kind of mathematical modeling methods such as the identification of linear system by input/output data and the regression analysis modeling method for prediction problem. PNN is an algorithm that obtains preferred network structure by generating consecutive layers as well as nodes by using a multivariate polynomial subexpression. It has much fewer nodes and more flexible adaptability than existing neural network algorithms. However, such algorithms lead to overfitting problems due to noise sensitivity as well as excessive trainning while generation of successive network layers. To alleviate such overfitting problem and also effectively design its ensuing deep network structure, two techniques are introduced. That is we use the two techniques of both SSC(Sum of Squared Coefficients) and $L_2$ regularization for consecutive generation of each layer's nodes as well as each layer in order to construct the deep PNN structure. The technique of $L_2$ regularization is used for the minimum coefficient estimation by adding penalty term to cost function. $L_2$ regularization is a kind of representative methods of reducing the influence of noise by flattening the solution space and also lessening coefficient size. The technique for the SSC is implemented for the minimization of Sum of Squared Coefficients of polynomial instead of using the square of errors. In the sequel, the overfitting problem of the deep PNN structure is stabilized by the proposed method. This study leads to the possibility of deep network structure design as well as big data processing and also the superiority of the network performance through experiments is shown.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.4
• /
• pp.534-540
• /
• 2014

In this paper, we propose a fuzzy combined Polynomial Neural Network(PNN) for pattern classification. The fuzzy combined PNN comes from the generic TSK fuzzy model with several linear polynomial as the consequent part and is the expanded version of the fuzzy model. The proposed pattern classifier has the polynomial neural networks as the consequent part, instead of the general linear polynomial. PNNs are implemented by stacking the simple polynomials dynamically. To implement one layer of PNNs, the various types of simple polynomials are used so that PNNs have flexibility and versatility. Although the structural complexity of the implemented PNNs is high, the PNNs become a high order-multi input polynomial finally. To estimate the coefficients of a polynomial neuron, The weighted linear discriminant analysis. The output of fuzzy rule system with PNNs as the consequent part is the linear combination of the output of several PNNs. To evaluate the classification ability of the proposed pattern classifier, we make some experiments with several machine learning data sets.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.3
• /
• pp.619-625
• /
• 2000

A many control techniques have been proposed in order to improve the control performance of the discrete-time domain control system. In the position control system, the output of a controller is generally used as the input of a plant but the undesired noise is included in the output of a controller. Therefore there is a need to used a precompensator for rejecting the undesired noise. In this paper, The expanded I-PD control system with a precompensator is constructed. The precompensator and I-PD controller are designed by a neural network and these coefficients are changed automatically to be a desired response of system when the response characteristic of system is changed under a condition.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.6A
• /
• pp.640-650
• /
• 2004

Channel estimation schemes are proposed for Multiple Input-Multiple output Orthogonal Frequency Division Multiplexing(MIMO-OFDM) systems based on the physical layer specification of the IEEE 802.1 la. By combining the space-time block coding(STBC)/ space-frequency block coding(SFBC) techniques with the transform domain interpolation, the proposed algorithms achieve more accurate channel coefficients for the MIMO channels such that improve the BER performance. The performance improvements of the proposed algorithms are evaluated by simulations under the various multipath fading channel environments and various transmission rates.

• Smart Structures and Systems
• /
• v.17 no.6
• /
• pp.1017-1030
• /
• 2016

This study focuses on the system identification of reinforced concrete bridges using vector autoregressive model (VAR). First, the time series output response from a bridge establishes the autoregressive (AR) models. AR models are one of the most accurate methods for stationary time series. Burg's algorithm estimates the autoregressive coefficients (ARCs) at p-lag by reducing the sum of the forward and the backward errors. The computed ARCs are assembled in the state system matrix and the eigen-system realization algorithm (ERA) computes: the eigenvector matrix that contains the vectors of the mode shapes, and the eigenvalue matrix that contains the associated natural frequencies. By taking advantage of the characteristic of the AR model with ERA (ARMERA), civil engineering can address problems related to damage detection. Operational modal analysis using ARMERA is applied to three experiments. One experiment is coupled with an artificial neural network algorithm and it can detect damage locations and extension. The neural network uses a specific number of ARCs as input and multiple submatrix scaling factors of the structural stiffness matrix as output to represent the damage.

• Steel and Composite Structures
• /
• v.46 no.6
• /
• pp.731-744
• /
• 2023

Vehicle load information is an important role in operating and ensuring the structural health of cable-stayed bridges. In this regard, an efficient and economic method is proposed for vehicle load detection based on the observed cable tension and vehicle position using a graph neural network (GNN). Datasets are first generated using the practical advanced analysis program (PAAP), a robust program for modeling and considering both geometric and material nonlinearities of bridge structures subjected to vehicle load with low computational costs. With the superiority of GNN, the proposed model is demonstrated to precisely capture complex nonlinear correlations between the input features and vehicle load in the output. Four popular machine learning methods including artificial neural network (ANN), decision tree (DT), random forest (RF), and support vector machines (SVM) are refereed in a comparison. A case study of a cable-stayed bridge with the typical truck is considered to evaluate the model's performance. The results demonstrate that the GNN-based model provides high accuracy and efficiency in prediction with satisfactory correlation coefficients, efficient determination values, and very small errors; and is a novel approach for vehicle load detection with the input data of the existing monitoring system.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.5
• /
• pp.1031-1039
• /
• 2015

This study examines the economic ripple effect on the domestic information security manufacturing and service sectors through input-output analysis. The production inducement coefficient of the manufacturing sector is bigger than the average of whole industry, but that of the service sector is smaller than the average. On the other hand, the service sector is superior to the manufacturing sector in the value added and employment inducement coefficients. Forward and backward linkage effects of manufacturing and service sectors are generally lower than those of the average of whole industry. The information security industry has insignificant contribution to national economy and employment growth overall. In particular, the manufacturing sector records minus contribution to employment growth, which means that a lot of effort for increasing employment must be given further on in the sector.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.1B
• /
• pp.89-98
• /
• 2000

This paper presents an active noise cancelling model using wavelet transform and subband filter banks based on adaptive filter. The analysis filter banks decompose input and error signals into QMF filter banks of lowpass and highpass bands. Each filter bank uses wavelet filter with dyadic tree structure. The decomposed input and error signals are iterated by adaptive filter coefficients of each subband using filtered-X LMS algorithm. The synthesis filter banks make output signal of wideband with perfect reconstruction to prepare adaptive filter output signals of each subband. The analysis and synthesis niter hants use conjugate quadrature filters for Pefect reconstruction. Also, The delayed LMS algorithm model for on-line identification of error path transfer characteristics is used gain and acoustic time delay factors. The proposed adaptive active noise cancelling modelis suggested by system retaining the computational and convergence speed advantage using wavelet subband filter banks.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.23 no.2
• /
• pp.125-130
• /
• 2009

With respect to the goal of achieving at least 50[%] reduction of global emissions by 2050, the G8 leaders agreed to seek to share and adopt it with all Parties to the UNFCCC in the 34th Group of Eight Summit held in Toyako, Japan in July 2008. Korea is also expected to obey the Kyoto Protocol starting in 2013, which will result in a serious shock especially to the electric power industry. The power plants burning the fossil fuel produce more than 20 percent of national $CO_2$ emission. This paper presents an economic feasibility study on fuel transfer for a thermal power plant considering $CO_2$ emission cost. Calculation of the breakeven point for the fuel transfer from LNG to heavy oil of D power plant is demonstrated using the input-output coefficients obtained by the performance test.