• Journal of the Korean Operations Research and Management Science Society
• /
• v.31 no.1
• /
• pp.15-24
• /
• 2006

In simulation input modeling, it is important to identify a probability distribution to represent the input process of interest. In this paper, an appropriate sample size is determined for parameter estimation associated with some typical probability distributions frequently encountered in simulation input modeling. For this purpose, a statistical measure is proposed to evaluate the effect of sample size on the precision as well as the accuracy related to the parameter estimation, square rooted mean square error to parameter ratio. Based on this evaluation measure, this sample size effect can be not only analyzed dimensionlessly against parameter's unit but also scaled regardless of parameter's magnitude. In the Monte Carlo simulation experiments, three continuous and one discrete probability distributions are investigated such as ; 1) exponential ; 2) gamma ; 3) normal ; and 4) poisson. The parameter's magnitudes tested are designed in order to represent distinct skewness respectively. Results show that ; 1) the evaluation measure drastically improves until the sample size approaches around 200 ; 2) up to the sample size about 400, the improvement continues but becomes ineffective ; and 3) plots of the evaluation measure have a similar plateau pattern beyond the sample size of 400. A case study with real datasets presents for verifying the experimental results.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.37 no.3
• /
• pp.180-185
• /
• 1988

This paper considers the problem of the optimal input design for parameter estimtion in the ARMAX model in which the system and noise transfer function have the common denominator polynomial. Deriving the information matrix, in detail, for the assumed model structure and using the autocorrelation functin of the filtered input as design variables, it is shown that D-optimal input signal can be realized as an autoregressive moving average process. Computer simulations are carried out to show the standard-deviation reduction in the parameter estimtes using the optimal input signal.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.38 no.2
• /
• pp.127-136
• /
• 1989

This paper describes the parameter convergence properties of an adaptive system to identify a single-input single-output plant model. It is demonstrated that, by using power spectrum analysis, the persistency of excitation (PE) condition in order to guarantee the exponential stability of the adaptive control system can be transformed into the positive definite behavior for the auto-correlation function matrix of adaptive signal. The existence of parameter nominal values can be analyzed by this condition and the convergence rates of parameter are determined by examining the auto-correlation function. We may use the sufficient richness (SR) of input spectrum instead of the PE condition to analyze the parameter boundedness. It can be shown that the eigen values of the auto-correlation function are always related with adaptive gain, input amplitude and positions or numbers of input spectra. In each case, the variation of parameter convergence rate can be also verified.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.12
• /
• pp.59-65
• /
• 2014

Two input resistances converted from $S_{11}$-parameter and $Z_{11}$-parameter of MOSFETs with various gate finger numbers Nf were measured in low frequency region. The 1/Nf dependent input resistance from $S_{11}$-parameter exhibits much lower values than that from $Z_{11}$-parameter in the range of $Nf{\leq}64$. This 1/Nf dependence was theoretically verified by using Nf dependent nonlinear equation derived from a MOSFET equivalent circuit.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.12
• /
• pp.15-19
• /
• 2016

The gate bias dependence of kink phenomenon with a large deviation from the resistance circle in Smith chart is observed in the frequency response of $S_{11}$-parameter for large multi-finger RF MOSFETs. For the first time, this bias dependence is analyzed by measuring magnitude and phase of $S_{11}$-parameter, input resistance and input capacitance. As a result, $V_{gs}$ dependent $S_{11}$-parameter is largely changed by the magnitude of input capacitance as well as dominant pole and zero frequencies of input resistance. At $V_{gs}=0V$, the kink phenomenon occurs in the high frequency region because of very small phase difference of $S_{11}$-parameter and high pole frequency of input resistance. However, the kink phenomenon at higher $V_{gs}$ is generated in the low frequency region owing to large phase difference and low pole frequency.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.42 no.2
• /
• pp.35-48
• /
• 2017

As the technology convergence paradigm emerges, the need for "CIA techniques" to analyze the mutual effects of technology is increasing. However, since the CIA input parameter estimation is difficult, the present study suggests a "CIA input parameter setting model" to alleviate the difficulty of CIA input parameter estimation. This paper is focused on the difference of measurement difficulty by each scale which expert's estimation behavior was defined as measurement activity quantifying the judgment of future technology. Therefore, this model is designed to estimate the input variable as a sequence or isometric scale that is relatively easy to measure, and then converts it into a probability value. The input parameter setting model of the CIA technique consists of three sub-models : 'probability value derivation model', 'influence estimation model', and 'impact value calculation model', in order to develop a series of models the Thurstone V model, Regression Analysis, etc has been used.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.48 no.4
• /
• pp.410-418
• /
• 1999

This paper describes the use of residual and parameter identification methods for fault detection in an air handling unit. Faults can be detected by comparing expected condition with the measured faulty data using residuals. Faults can also be detected by examining unmeasurable parameter changes in a model of a controlled system using a system identification technique. In this study, AutoRegressive Moving Average with seXtrnal input(ARMAX) and AutoRegressive with eXternal input(ARX) models with both single-input/single-input and multi-input/single-input structures are examined. Model parameters are determined using the Kalman filter recursive identification method. Regression equations are calculated from normal experimental data and are used to compute expected operating variables. These approaches are tested using experimental data from a laboratory's variable-air-volume air-handling-unit.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10b
• /
• pp.1655-1658
• /
• 1991

This paper considers the design and analysis of one-step ahead optimal and adaptive controllers, under the restriction that a known constraint on the input amplitude is imposed. It is assumed that the discrete-time single-input, single-output system to be controlled is linear, except for inequality constraints on the input. The objective function to be minimized is an one-step quadratic function, where polynomial weights on the input and output are included. Both the known parameter and unknown parameter (indirect adaptive controller) cases are examined.

• Journal of Korean Institute of Industrial Engineers
• /
• v.27 no.1
• /
• pp.47-53
• /
• 2001

In this paper, we compare the classification performances of both ensemble and clustering algorithms (Data Bagging, Variable Selection Bagging, Parameter Combining, Clustering) to logistic regression in consideration of various characteristics of input data. Four factors used to simulate the logistic model are (1) correlation among input variables (2) variance of observation (3) training data size and (4) input-output function. In view of the unknown relationship between input and output function, we use a Taguchi design to improve the practicality of our study results by letting it as a noise factor. Experimental study results indicate the following: When the level of the variance is medium, Bagging & Parameter Combining performs worse than Logistic Regression, Variable Selection Bagging and Clustering. However, classification performances of Logistic Regression, Variable Selection Bagging, Bagging and Clustering are not significantly different when the variance of input data is either small or large. When there is strong correlation in input variables, Variable Selection Bagging outperforms both Logistic Regression and Parameter combining. In general, Parameter Combining algorithm appears to be the worst at our disappointment.

• Proceedings of the IEEK Conference
• /
• 2000.11b
• /
• pp.245-248
• /
• 2000

In this paper, we present a high-performance SiGe HBT's RF input impedance parameter extraction method. The SiGe HBT has emitter width of 0.5${\mu}{\textrm}{m}$ and length of 6${\mu}{\textrm}{m}$. S-parameter has been measured with the collector current of 1~3㎃ using on-wafer RF measuring system . The pre-calculation method was used in order to overcome the local minimum problem. This method enabled us to extract a RF(1~10㎓) input impedance parameter.