• Ecology and Resilient Infrastructure
• /
• v.7 no.2
• /
• pp.114-125
• /
• 2020

Despite the various artificial neural networks that have been developed, most of the discharge models in previous studies have been developed using deep neural networks. This study aimed to develop a discharge model using a convolution neural network (CNN), which was used to solve classification problems. Furthermore, the applicability of CNN was evaluated. The photographs (pictures or images) for input data to CNN could not clearly show the characteristics of the study area as well as precipitation. Hence, the model employed in this study had to use numerical images. To solve the problem, the CN of NRCS was used to generate images as input data for the model. The generated images showed a good possibility of applicability as input data. Moreover, a new application of CN, which had been used only for discharge prediction, was proposed in this study. As a result of CNN training, the model was trained and generalized stably. Comparison between the actual and predicted values had an R² of 0.79, which was relatively high. The model showed good performance in terms of the Pearson correlation coefficient (0.84), the Nash-Sutcliffe efficiency (NSE) (0.63), and the root mean square error (24.54 ㎥/s).

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.6
• /
• pp.563-567
• /
• 2001

We generalize the widebane P0-weyl symbol (P0WS) and the widebane spreading function (WSF) using the generalized warping function . The new generalized P0WS and WSF are useful for analyzing system and communication channels producing generalized time shifts. We also investigated the relationship between the affine Wey1 symbol(AWS) and the P0WS. By using specific warping functions, we derive new P0WS and WSF as analysis tools for systems and communication channels with non-linear group delary characteristics. The new P0WS preserves specific types of changes imposed on random processes. The new WSF provides a new interpretation of output of system and communication channel as weighted superpositions of non-linear time shifts on the input. It is compared to the conventional method obtaining output of system and communication channel as a convention integration of the input with the impulse response of the system and the communication channel. The convolution integration can be interpreted as weighted superpositions of liner time shifts on the input where the weight is the impulse response of the system and the communication channel. Application examples in analysis and detection demonstrate the advantages of our new results.

• International Journal of Aeronautical and Space Sciences
• /
• v.13 no.4
• /
• pp.458-467
• /
• 2012

In this study, the linear viscoelastic response of a rectangular laminated plate is investigated. The viscoelastic properties, expressed by two basic spring-dashpot models, that is Kelvin and Maxwell models, is assumed in the range to investigate the influence of viscoelastic coefficients to mechanical behavior. In the present study, viscoelastic responses are performed for two popular equivalent single-layered theories, such as the first-order shear deformation theory (FSDT) and third-order shear deformation theory (TSDT). Compliance and relaxation modulus of time-dependent viscoelastic behavior are approximately determined by Prony series. The constitutive equation for linear viscoelastic material as the Boltzmann superposition integral equation is simplified by the convolution theorem of Laplace transformation to avoid direct time integration as well as to improve both accuracy and computational efficiency. The viscoelastic responses of composite laminates in the real time domain are obtained by applying the inverse Laplace transformation. The numerical results of viscoelastic phenomena such as creep, cyclic creep and recovery creep are presented.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.2
• /
• pp.29-35
• /
• 1998

This paper describes architectures and design of a SIMD type parallel image processing chip called SliM-II. The chiphas a linear array of 64 processing elements (PEs), operates at 30 MHz in the worst case simulation and gives at least 1.92 GIPS. In contrast to existing array processors, such as IMAP, MGAP-2, VIP, etc., each PE has a multiplier that is quite effective for convolution, template matching, etc. The instruction set can execute an ALU operation, data I/O, and inter-PE communication simulataneously in a single instruction cycle. In addition, during the ALU/multiplier operation, SliM-II provides parallel move between the register file and on-chip memory as in DSP chips, SliM-II can greatly reduce the inter-PE communication overhead, due to the idea a sliding, which is a technique of overlapping inter-PE communication with computation. Moreover, the bandwidth of data I/O and inter-PE communication increases due to bit-parallel data paths. We used the COMPASS$^{TM}$ 3.3 V 0.6.$\mu$m standrd cell library (v8r4.10). The total number of transistors is about 1.5 muillions, the core size is 13.2 * 13.0 mm$^{2}$ and the package type is 208 pin PQ2 (Power Quad 2). The performance evaluation shows that, compared to a existing array processors, a proposed architeture gives a significant improvement for algorithms requiring multiplications.s.

• Bulletin of the Society of Naval Architects of Korea
• /
• v.24 no.4
• /
• pp.9-18
• /
• 1987

The hydrodynamic radiation forces acting on a ship travelling in waves have been conventionally treated by strip theories or by direct three dimensional approaches, most of which have been formulated in frequency domain. If the forward speed of a ship varies with time, or if its path is not a straight line, conventional frequency domain analysis can no more be used, and for these cases time domain analysis may be used. In this paper, formulations are made in time domain with applications to some problems the results of which are known in frequency domain. And the results of both domains are compared to show the characteristics and validity of time domain solutions. The radiation forces acting on a three dimensional body within the framework of a linear theory. If the linearity of entire system is assumed, radiation forces due to arbitrary ship motions can be expressed by the convolution integral of the arbitrary motion velocity and the so called impulse response function. Numerical calculations are done for some bodies of simple shapes and Series-60[$C_B=0.7$] ship model. For all cases, integral equation techniques with transient Green's function are used, and velocity or acceleration potentials are obtained as the solution of the integral equations. In liner systems, time domain solutions are related with frequency domain solutions by Fourier transform. Therefore time domain solutions are Fourier transformed by suitable relations and the results are compared with various frequency domain solutions, which show good agreements.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.5B
• /
• pp.859-870
• /
• 2000

In this paper, an efficient H/W implementation technique for guard interval in OFDM(Orthogonal Frequency Division Multiplexing) systems is proposed and applied to ASIC chip design of an OFDM-based 25 Mbps wireless ATM modem. In OFDM systems, a cyclic prefix, longer than the largest multipath delay spread, is usually inserted to maintain the orthogonality of subchannels, by making the linear convolution of the channel ok like circular convolution inherent to the discreate Fourier domain, as well as to prevent the ISI(Intersymbol Interference) within the OFDM block. However, the OFDM system using the cyclic prefix requires an additional H/W in transmitter in order to store the original samples and to append the cyclic prefix to the beginning of each block. In this paper, a new approach using a cyclic prefix, even with a significantly lower H/W complexity. Finally, the performance of the proposed approach is demonstrated by applying it to ASIC chip design of an OFDM-based 25 Mbps wireless ATM modem.

• Journal of the Society of Naval Architects of Korea
• /
• v.30 no.2
• /
• pp.43-53
• /
• 1993

There exist two difficulties in the nonlinear wave-body problems. First is the abrupt behavior near the intersection point between the body and the free surface, and second is the far field treatment. In this paper, the far field treatment is considered. The main idea is the Taylor series expansion of free-surface geometry and the application of F.F.T. algorithm. The numerical step is as follows. The velocity potential is expressed by the Green's theorem. and the solution is obtained by iteration method. In the iteration stage, the expressions by the Green's theorem are transformed to the convolution forts with the expansion of free surface by the wave slope. Here F.F.T. is applied, so the computing time can be of O(Nlog N) where N is the number of unknowns. The numerical analysis is carried out and the results are compared with other results in linear floating body problem and nonlinear moving pressure patch problem, and good agreements are obtained. Finally nonlinear floating body radiation problem is carried out with computing time of O(Nlog N).

• Journal of the Korea Society for Simulation
• /
• v.30 no.4
• /
• pp.1-8
• /
• 2021

Deep learning is one of the most widely accepted methods for the forecasting of time series data which have the complexity and non-linear behavior. In this paper, we investigate the modification of a state-of-art WaveNet deep learning architecture and walk forward validation (WFV) in order to forecast electric power consumption data 24-hour-ahead. WaveNet originally designed for raw audio uses 1D dilated causal convolution for long-term information. First of all, we propose a modified version of WaveNet which activates real numbers instead of coded integers. Second, this paper provides with the training process with tuning of major hyper-parameters (i.e., input length, batch size, number of WaveNet blocks, dilation rates, and learning rate scheduler). Finally, performance evaluation results show that the prediction methodology based on WFV performs better than on the traditional holdout validation.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.11 no.1
• /
• pp.19-24
• /
• 2011

Edges represent significant boundary information between objects or classes. Various methods, which are based on differential operation, such as Sobel, Prewitt, Roberts, Canny, and etc. have been proposed and widely used. The methods are based on a linear convolution of mask with pre-assigned coefficients. In this paper, we propose an edge detection method based on Dempster-Shafer's evidence theory to evaluate edgeness of the given pixel. The effectiveness of the proposed method is shown through experimental results on several test images and compared with conventional methods.

• Bulletin of the Korean Mathematical Society
• /
• v.54 no.6
• /
• pp.2065-2079
• /
• 2017

This paper is devoted to the minimal and maximal bounded solutions for general time interval quadratic backward stochastic differential equations with stochastic conditions. A general existence result is established by the method of convolution, the exponential transform, Girsanov's transform and a priori estimates, where the terminal time is allowed to be finite or infinite, and the generator g is allowed to have a stochastic semi-linear growth and a general growth in y, and a quadratic growth in z. This improves some existing results at some extent. Some new ideas and techniques are also applied to prove it.