• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.12
• /
• pp.1295-1304
• /
• 2004

In this paper, the Bayesian recurrent neural network is proposed to predict time series data. A neural network predictor requests proper learning strategy to adjust the network weights, and one needs to prepare for non-linear and non-stationary evolution of network weights. The Bayesian neural network in this paper estimates not the single set of weights but the probability distributions of weights. In other words, the weights vector is set as a state vector of state space method, and its probability distributions are estimated in accordance with the particle filtering process. This approach makes it possible to obtain more exact estimation of the weights. In the aspect of network architecture, it is known that the recurrent feedback structure is superior to the feedforward structure for the problem of time series prediction. Therefore, the recurrent neural network with Bayesian inference, what we call Bayesian recurrent neural network (BRNN), is expected to show higher performance than the normal neural network. To verify the proposed method, the time series data are numerically generated and various kinds of neural network predictor are applied on it in order to be compared. As a result, feedback structure and Bayesian learning are better than feedforward structure and backpropagation learning, respectively. Consequently, it is verified that the Bayesian reccurent neural network shows better a prediction result than the common Bayesian neural network.

• Communications for Statistical Applications and Methods
• /
• v.24 no.1
• /
• pp.1-13
• /
• 2017

This manuscript summarized advances in bootstrap methods for long-range dependent time series data. The stationary linear long-memory process is briefly described, which is a target process for bootstrap methodologies on time-domain and frequency-domain in this review. We illustrate time-domain bootstrap under long-range dependence, moving or non-overlapping block bootstraps, and the autoregressive-sieve bootstrap. In particular, block bootstrap methodologies need an adjustment factor for the distribution estimation of the sample mean in contrast to applications to weak dependent time processes. However, the autoregressive-sieve bootstrap does not need any other modification for application to long-memory. The frequency domain bootstrap for Whittle estimation is provided using parametric spectral density estimates because there is no current nonparametric spectral density estimation method using a kernel function for the linear long-range dependent time process.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.8
• /
• pp.1647-1652
• /
• 2009

Generally the neural network and the fuzzy compensative algorithm are applied to forecast the time series for power demand with a characteristic of non-linear dynamic system, but it has a few prediction errors relatively. It also makes long term forecast difficult for sensitivity on the initial condition. On this paper, we evaluate the chaotic characteristic of electrical power demand with analysis methods of qualitative and quantitative and perform a forecast simulation of electrical power demand in regular sequence, attractor reconstruction, time series forecast for multi dimension using Lyapunov exponent quantitatively. We compare simulated results with the previous method and verify that the purpose one being more practice and effective than it.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.05a
• /
• pp.171-174
• /
• 2009

Generally the neural network and the fuzzy compensative algorithm are applied to forecast the time series for power demand with a characteristic of non-linear dynamic system, but it has a few prediction errors relatively. It also makes long term forecast difficult for sensitivity on the initial condition. On this paper, we evaluate the chaotic characteristic of electrical power demand with analysis methods of qualitative and quantitative and perform a forecast simulation of electrical power demand in regular sequence, attractor reconstruction, time series forecast for multi dimension using Lyapunov exponent quantitatively. We compare simulated results with the previous method and verify that the purpose one being more practice and effective than it.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.5
• /
• pp.1312-1322
• /
• 1999

The world of linearity, which is regular, predictable and irrelevant to time sequence in most natural phenomenon, is a very small part. In fact, signals generated from natural phenomenon with which we're in contact are showed only slight linearity. Therefore it is very difficult to understand and analyze natural phenomenon with only predictable and regular linear systems. Due to these reasons researches concerning non-linear signals that of analysis were excluded being regarded as noise are being actively carried out. Countless signals generated from nonlinear system have the information about itself, and analyzing those signals and get information from it, that will be able to be used effectively in so may fields. Hence, in this paper we used a higher order spectrum, especially the bispectrum. After we prove the validity applying bispectrum to logistic map, which is typical chaotic signal. Subsequently by showing the result applying for actual signal analysis of EEG according to auditory stimuli, we show that higher order spectra is a very useful parameter in analysis of non-linear signals and the result of EEG analysis according to auditory stimuli.

• Earthquakes and Structures
• /
• v.4 no.1
• /
• pp.109-132
• /
• 2013

Current codes incorporate simplified methods for the prediction of acceleration demands on secondary structural and non-structural elements at different levels of a building. While the use of simple analysis methods should be advocated, damage to both secondary structural and non-structural elements in recent earthquakes have highlighted the need for improved design procedures for such elements. In order to take a step towards the formation of accurate but simplified methods of predicting floor spectra, this work examines the floor spectra on elastic and inelastic single-degree of freedom systems subject to accelerograms of varying seismic intensity. After identifying the factors that appear to affect the shape and intensity of acceleration demands on secondary structural and non-structural elements, a new series of calibrated equations are proposed to predict floor spectra on single degree of freedom supporting structures. The approach uses concepts of dynamics and inelasticity to define the shape and intensity of the floor spectra at different levels of damping. The results of non-linear time-history analyses of a series of single-degree of freedom supporting structures indicate that the new methodology is very promising. Future research will aim to extend the methodology to multi-degree of freedom supporting structures and run additional verification studies.

• International Journal of Computer Science & Network Security
• /
• v.23 no.4
• /
• pp.69-78
• /
• 2023

With the global rise of digital data, the uncontrolled quantity of data is susceptible to cyber warfare or cyber attacks. Therefore, it is necessary to improve cyber security systems. This research studies the behavior of malicious acts and uses Higuchi Fractal Dimension (HFD), which is a non-linear mathematical method to examine the intricacy of the behavior of these malicious acts and anomalies within the cyber physical system. The HFD algorithm was tested successfully using synthetic time series network data and validated on real-time network data, producing accurate results. It was found that the highest fractal dimension value was computed from the DoS attack time series data. Furthermore, the difference in the HFD values between the DoS attack data and the normal traffic data was the highest. The malicious network data and the non-malicious network data were successfully classified using the Receiver Operating Characteristics (ROC) method in conjunction with a scaling stationary index that helps to boost the ROC technique in classifying normal and malicious traffic. Hence, the suggested methodology may be utilized to rapidly detect the existence of abnormalities in traffic with the aim of further using other methods of cyber-attack detection.

• Proceedings of the IEEK Conference
• /
• 2000.07b
• /
• pp.1100-1102
• /
• 2000

Most systems in tile real world are non-linear and can be represented by the non-linear autoregressive moving average (NARMA) model. The extension of fuzzy system for modeling the system that is represented by NARMA model will be proposed in this paper. Here, fuzzy basis function (FBF) is used as fuzzy NARMA(p,q) model. Then, an approach to Identify fuzzy NARMA models based on fuzzy basis functions is proposed. The efficacy of the proposed approach is shown from experimental results.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.1
• /
• pp.151-160
• /
• 2017

This paper presents a new technique of reference current generator based on Kalman filter (KF) estimator for three-phase shunt active power filter (APF). The stationary reference frame (d-q algorithm) is used to transform the load currents into DC component. The harmonics of load currents are extracted and the three-phase reference currents are generated using KF estimator. The work is simulated using Matlab/Simulink platform. To validate the simulation results, an experimental test-rig have been perform using real-time control dSPACE DS1104. In addition, hysteresis current control was used to generate the switching signal for the correction of the harmonics in the system. The non-linear load were constructed with three-phase rectifier which connected in series with inductor and parallel with resistor and capacitor. The results shows that the new technique of shunt APF embedded with KF is proven to eliminate the harmonics created by the non-linear load with some improvement on the total harmonics distortion (THD).

• Journal of the Society of Naval Architects of Korea
• /
• v.42 no.1 s.139
• /
• pp.43-49
• /
• 2005

in this paper a dolphin-fender moored pontoon-type floating structure in shallow water depth is studied focusing on mooring force. The pontoon-type floating structure is 500m long, 300m wide. The structure has partially non-uniform drafts of 2.0m and 3.0m. The employed mooring system is a guyed frame type dolphin-fender system. The 1/125 scale model fender system is made of rubber tube to have hi-linear load deflection characteristics. A series of model tests has been conducted focusing on motion and fender force responses in regular and irregular waves at KRISO's ocean engineering basin Non-linear numerical simulation of fender reaction force has been carried out and the results are compared with those of model tests. The simulated rigid body motion and mooring forces also have been compared with the test results.