• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10b
• /
• pp.1323-1326
• /
• 2001

서브네트워크에서 실시간으로 통신 트래픽을 감시하고, 트래픽 정보를 바탕으로 시계열 분석을 이용해 트래픽의 변화추이를 예측할 수 있는 시스템을 설계 및 구현한다. SNMP를 이용한 MIB-II 정보를 바탕으로 하는 분석 방법은 누적 데이터를 기본으로 하는 관리 방법으로 이상 징후의 판단이 실시간 감시에는 적합하지 않은 점이 있다. 따라서, 본 논문에서는 실시간 트래픽 감시를 위해 서브네트워크에 들어오거나 나가는 트래픽의 양을 측정하여 분석하고, 이 정보를 바탕으로 특정 시점 이후의 트래픽 추이를 시계열 분석 방법을 이용하여 미래의 트래픽 양을 예측하는 알고리즘을 시스템으로 구현한다. 예측 알고리즘으로는 AR, MA, ARMA, ARIMA 모델중에 평균 제곱 오차를 최소로 가지는 알고리즘을 선택하여 예측하도록 설계한다. 개발되는 시스템을 망 관리자가 전체 통신 네트워크의 부하 상태를 예상할 수 있게 하여 신속하고 예방적인 대응을 할 수 있다.

• Proceedings of the Korean Statistical Society Conference
• /
• 2002.11a
• /
• pp.265-272
• /
• 2002

An Integer-valued autoregressive integrated (INARI) model is introduced to eliminate stochastic trend and seasonality from time series of count data. This INARI extends the previous integer-valued ARMA model. We show that it is stationary and ergodic to establish asymptotic normality for conditional least squares estimator. Optimal estimating equations are used to reflect categorical and serial correlations arising from panel count data and variations arising from three random processes for obtaining observation into estimation. Under regularity conditions for martingale sequence, we show asymptotic normality for estimators from the estimating equations. Using cancer mortality data provided by the U.S. National Center for Health Statistics (NCHS), we apply our results to estimate the probability of cells classified by 4 causes of death and 6 age groups and to forecast death count of each cell. We also investigate impact of three random processes on estimation.

• Journal of Korean Institute of Industrial Engineers
• /
• v.38 no.1
• /
• pp.17-24
• /
• 2012

We address some potential problems with the existing procedures of outlier detection in time series. Also we propose modifications in estimating model parameters and outlier effects in order to reduce the number of tests and to increase the detection accuracy. Experiments with some artificial data sets show that the proposed procedure significantly reduces the number of tests and enhances the accuracy of estimated parameters as well as the detection power.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1993.04b
• /
• pp.86-91
• /
• 1993

연삭숫돌은 드레싱에 의해 초기숫돌면의 상태가조정되며, 가공과 더불어 공작물과 연삭숫돌의 연속적인 간섭으로 공작물의 치수변화와함께 연삭숫돌도 작업면이 변화하게 된다. 공작물의 표면은 연삭조건 및 공작물과 숫돌의 접촉특성에 따라 Rubbging, Ploughing, Cutting 작용이 이루어져 가공면이 형성되며 이때의 공작물의 표면 품위( Surface Integrity)는 연삭숫돌의 상태변화에따라 매우 민감하게 영향을 받는다. 본 연구에서는 원통연삭작업 에서의 감시 및 진단기술개발을 위하여 AE 센서의 적용가능성을 살펴보기위하여, 센서의 설치위치에 따른 영향 및 연삭가공 특성을 잘나타내는 AE Parameter의 선정을 위한 실험을 수행하며, 가공시간에 따른 연삭숫돌작업면 의 상태변화에 의한 AE Parameter의 변화를 조사하며, 이와더불어 연삭중의 법선저항을 측정분석하고 통계적인 수법으로 ARMA (Autoregressive Moving Average)을 이용하여 가공특성변화를 분석한다.

• Journal of Intelligence and Information Systems
• /
• v.3 no.1
• /
• pp.47-68
• /
• 1997

This paper discusses a method of modified Elman network(1990) for nonlinear predictions and its a, pp.ication to forecasting daily exchange rate returns. The method consists of two stages that take advantages of both time domain filter and modified feedback networks. The first stage straightforwardly employs the filtering technique to remove extreme noise. In the second stage neural networks are designed to take the feedback from both hidden-layer units and the deviation of outputs from target values during learning. This combined feedback can be exploited to transfer unconsidered information on errors into the network system and, consequently, would improve predictions. The method a, pp.ars to dominate linear ARMA models and standard dynamic neural networks in one-step-ahead forecasting exchange rate returns.

• Proceedings of the KIEE Conference
• /
• 1987.07b
• /
• pp.1422-1426
• /
• 1987

In this paper, an adaptive DPCM scheme is presented for encoding monochrome images with easy hardware implementation at a transmission rate of exactly 1 bit/pel. The system is mainly composed of a compensated mean predictor and an adaptive two-level quantizer with backward estimation. In this system, the predictor is a sort of two-dimensional ARMA predictor in which a moving-average part is added to the conventional mean predictor. The quantizer adapts to the local statistics of its input without overhead information. To reduce annoying granular noise in the reconstructed image, Lee filter is used after reconstruction in the receiver.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.306-308
• /
• 2006

산업현장에서 가장 많이 활용되고 있는 PID제어기의 이론적인 배경을 토대로 실제 DCS 기반 플랜트에서 PID Tuning Method에 의한 PID제어기를 구현하고 제어성능을 확인한다. 또한 DCS(Distribute Control System)의 PID Controller를 분석하고 전 공정제어 System 중 일부분을 ARMA Modeling하여 만족스런 성능이 구현되도록 최적의 PID gain Parameter를 찾는다.

• Proceedings of the KIPE Conference
• /
• 2012.11a
• /
• pp.33-34
• /
• 2012

As the distributed power system with PV and ESS is highlighted to be one of the most prominent structure to replace the traditional electric power system, power flow scheduling is expected to bring better system efficiency. Optimal energy management system (EMS) where the power from PV and the grid is managed in time-domain using ESS needs an optimization process. In this paper, main optimization method is implemented using dynamic programming (DP). To overcome the drawback of DP in which ideal future information is required, prediction stage precedes every EMS execution. A simple auto-regressive moving-average (ARMA) forecasting followed by a PI-controller updates the prediction data. Assessment of the on-line optimal EMS scheme has been evaluated on several cases.

• Journal of the Korea Society for Simulation
• /
• v.5 no.1
• /
• pp.1-12
• /
• 1996

We propose a new approach to classifying a time series data into one of the autoregressive moving-average (ARMA) models. It is bases on two pattern recognition concepts for solving time series identification. The one is an extended sample autocorrelation function (ESACF). The other is a neural network-driven decision tree classifier(NNDTC) in which two pattern recognition techniques are tightly coupled : neural network and decision tree classfier. NNDTc consists of a set of nodes at which neural network-driven decision making is made whether the connecting subtrees should be pruned or not. Therefore, time series identification problem can be stated as solving a set of local decisions at nodes. The decision values of the nodes are provided by neural network functions attached to the corresponding nodes. Experimental results with a set of test data and real time series data show that the proposed approach can efficiently identify the time seires patterns with high precision compared to the previous approaches.

• Journal of the Korea Society for Simulation
• /
• v.5 no.2
• /
• pp.59-72
• /
• 1996

The purpose of this study is to make an improvement to the batch-means method, which is a procedure to construct a confidence interval(c.i.) for the steady-state process mean of a stationary simulation output process. In the batch-means method, the data in the output process are grouped into batches. The sequence of means of the data included in individual batches is called a batch-menas process and can be treated as an independently and identically distributed set of variables if each batch includes sufficiently large number of observations. The traditional batch-means method, therefore, uses a batch size as large as possible in order to. destroy the autocovariance remaining in the batch-means process. The c.i. prodedure developed and empirically tested in this study uses a small batch size which can be well fitted by a simple ARMA model, and then utilizes the dependence structure in the fitted model to correct for bias in the variance estimator of the sample mean.