• Speech Sciences
• /
• v.14 no.1
• /
• pp.151-162
• /
• 2007

This study focuses on the classification of pathological voice using GMM (Gaussian Mixture Model) and compares the results to the previous work which was done by ANN (Artificial Neural Network). Speech data from normal people and patients were collected, then diagnosed and classified into two different categories. Six characteristic parameters (Jitter, Shimmer, NHR, SPI, APQ and RAP) were chosen. Then the classification method based on the artificial neural network and Gaussian mixture method was employed to discriminate the data into normal and pathological speech. The GMM method attained 98.4% average correct classification rate with training data and 95.2% average correct classification rate with test data. The different mixture number (3 to 15) of GMM was used in order to obtain an optimal condition for classification. We also compared the average classification rate based on GMM, ANN and HMM. The proper number of mixtures on Gaussian model needs to be investigated in our future work.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 2004.11a
• /
• pp.331-338
• /
• 2004

We present a three-dimensional average flow model considering elastic deformation of pad asperities for chemical mechanical planarization. To consider the contact deformation of pad asperities in the calculation of the flow factor, three-dimensional contact analysis of a semi-infinite solid based on the use of influence functions is conducted from computer generated three dimensional roughness data. The average Reynolds equation and the boundary condition of both force and momentum balance are used to investigate the effect of pad roughness and external pressure conditions on film thickness and wafer position angle.

• Journal of the Korean Solar Energy Society
• /
• v.30 no.4
• /
• pp.15-21
• /
• 2010

Radiation data are the best source of information for estimating average incident radiation. Lacking this or data from nearby locations of similar climate, it is possible to use empirical relation ships to estimate radiation from days of cloudiness. It is necessary to estimate the regression coefficients in order to predict the daily global radiation on a horizontal surface. There fore many different equations have proposed to evaluate them for certain areas. In this work a new correlation has been made to predict the solar radiation for 16 different areas over Korea by estimating the regression coefficients taking into account cloud cover. Particularly, the straight line regression model proposed shows reliable results for estimating the global radiation on a horizontal surface with monthly average deviation of -0.26 to +0.53% and each station annual average deviation of -1.61 to +1.7% from measured values.

• Proceedings of the KOR-KST Conference
• /
• 1995.02a
• /
• pp.84-98
• /
• 1995

Network-based model were developed to predict short term future traffic volume based on current traffic, historical average, and upstream traffic. It is presumed that upstream traffic volume can be used to predict the downstream traffic in a specific time period. Three models were developed for traffic flow prediction; a combination of historical average and upstream traffic, a combination of current traffic and upstream traffic, and a combination of all three variables. The three models were evaluated using regression analysis. The third model is found to provide the best prediction for the analyzed data. In order to balance the variables appropriately according to the present traffic condition, a heuristic adaptive weighting system is devised based on the relationships between the beginning period of prediction and the previous periods. The developed models were applied to 15-minute freeway data obtained by regular induction loop detectors. The prediction models were shown to be capable of producing reliable and accurate forecasts under congested traffic condition. The prediction systems perform better in the 15-minute range than in the ranges of 30-to 45-minute. It is also found that the combined models usually produce more consistent forecasts than the historical average.

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.329-330
• /
• 2015

From the results of previous my paper [10] in 2015 year of economic and electrical power storage research conference in Naju, this paper describes an application of autoregressive and moving average (ARMA) model to forecast hourly average wind speed (HAWS) in Jeju island. The models are used to build up short-term forecast of hourly average wind speed by the weighted sum of previous wind speed values.

• Proceedings of the KIPE Conference
• /
• 1998.10a
• /
• pp.609-614
• /
• 1998

A new small signal modeling of an average current mode control is proposed. In order to analyze the characteristics of the control scheme, the discrete and continuous time small signal models are derived. The derivation are mainly come from the analysis of the sampling effect presented in the current control loop. By the mathematical interpretation of practical sampler representing the sampling effect of a current control loop, the small signal models of an average current mode control can be easily derived. The instability of the current control loop, which gives rise to the subharmonic oscillation, can be identified by the proposed models. To show the usefulness of the proposed models, the simulation and experiment are carried out. The results show that the predicted results by the proposed model are much better agreed with the measured ones than that of the conventional model, even though the high gain of the compensation network of a current control loop is employed.

• Journal of Sensor Science and Technology
• /
• v.25 no.6
• /
• pp.418-422
• /
• 2016

A filtering algorithm based on the decentralized moving average Kalman filter with uncertainties is proposed in this paper. The proposed filtering algorithm presented combines the Kalman filter with the moving average strategy. A decentralized fusion algorithm with the weighted sum structure is applied to the local moving average Kalman filters (LMAKFs) of different window lengths. The proposed algorithm has a parallel structure and allows parallel processing of observations. Hence, it is more reliable than the centralized algorithm when some sensors become faulty. Moreover, the choice of the moving average strategy makes the proposed algorithm robust against linear discrete-time dynamic model uncertainties. The derivation of the error cross-covariances between the LMAKFs is the key idea of studied. The application of the proposed decentralized fusion filter to dynamic systems within a multisensor environment demonstrates its high accuracy and computational efficiency.

• The Journal of Fisheries Business Administration
• /
• v.46 no.2
• /
• pp.59-74
• /
• 2015

This study aims to estimate optimal harvesting production, fishing efforts, and stock levels of yellow croaker caught by the offshore Stow Net and the offshore Gill Net fisheries using the current value Hamiltonian method and the surplus production model. As analyzing processes, firstly, this study uses the Gavaris general linear model to estimate standardized fishing efforts of yellow croaker caught by the above multiple fisheries. Secondly, this study applies the Clarke Yoshimoto Pooley(CY&P) model among the various exponential growth models to estimate intrinsic growth rate(r), environmental carrying capacity(K), and catchability coefficient(q) of yellow croaker which inhabits in offshore area of Korea. Thirdly, the study determines optimal harvesting production, fishing efforts, and stock levels of yellow croaker using the current value Hamiltonian method which is including average landing price of yellow croaker, average unit cost of fishing efforts, and social discount rate based on standard of the Korean Development Institute. Finally, this study tries sensitivity analysis to understand changes in optimal harvesting production, fishing efforts, and stock levels of yellow croaker caused by changes in economic and biological parameters. As results drawn by the current value Hamiltonian model, the optimal harvesting production, fishing efforts, and stock levels of yellow croaker caught by the multiple fisheries were estimated as 19,173 ton, 101,644 horse power, and 146,144 ton respectively. In addition, as results of sensitivity analysis, firstly, if the social discount rate and the average landing price of yellow croaker continuously increase, the optimal harvesting production of yellow croaker increases at decreasing rate and then finally slightly decreases due to decreases in stock levels of yellow croaker. Secondly, if the average unit cost of fishing efforts continuously increases, the optimal fishing efforts of the multiple fisheries decreases, but the optimal stock level of yellow croaker increases. The optimal harvest starts climbing and then continuously decreases due to increases in the average unit cost. Thirdly, when the intrinsic growth rate of yellow croaker increases, the optimal harvest, fishing efforts, and stock level all continuously increase. In conclusion, this study suggests that the optimal harvesting production and fishing efforts were much less than actual harvesting production(35,279 ton) and estimated standardized fishing efforts(175,512 horse power) in 2013. This result implies that yellow croaker has been overfished due to excessive fishing efforts. Efficient management and conservative policy on stock of yellow croaker need to be urgently implemented.

• Journal of Preventive Medicine and Public Health
• /
• v.11 no.1
• /
• pp.98-109
• /
• 1978

The study was carried out for 101 establishments in Kyong-In areas to evaluate the industrial health management status utilizing the scoring method by Webb's model during the period from 1 September to October 30, 1977. To compare the results, reevaluation was made with the 45 questions' model prepared by 6 Korean professionals who were specialized in industrial health fields. The results were as follows: 1. The mean rate of affirmative answers for 101 establishments was 51.6%. The mean weighted score rate of affirmative answers was 52.3%. 2. The mean rate of affirmative answers on components of the philosophy and facility resource for 101 establishments was higher than that of average rate. The mean rate of affirmative answers on components of the health evaluation and health management among the health service program was lower than that of average rate. 3. The mean rate of affirmative answers on components was highter among the establishments with more than 500 employees. The mean rate of affirmative answers of chemical establishment was lower than that of others. 4. The mean rate of affirmative answers on 45 questions' model for 101 establishments were 67.1%. The mean weighted score rate of affirmative answers was 70.0%. 5. In case of 45 questions' model, the mean rate of affirmative answers on components of philosophy and treatment was higher than that of average rate and the mean rate of affirmative answers on components of the facility resource and the health evaluation was lower than that of average rate. 6. The mean rate of affirmative answers of the 45 Questions' model was higher than that of Webb's model in size and class of 101 establishments. Author concluded that Webb's model must be suitable for evaluating higher conditions of occupational health management than is presently used in Korean establishments. According to the results, however, there were no significant differences between Webb's model and the 45 questions' model. So it could be used to evaluate Occupational Health Program. For this objective, in Korean occupational situations, further study also must be made comprehensively thereafter.

• Wind and Structures
• /
• v.36 no.4
• /
• pp.237-247
• /
• 2023

The emergence of high-rise buildings has necessitated frequent structural health monitoring and maintenance for safety reasons. Wind causes damage and structural changes on tall structures; thus, safe structures should be designed. The pressure developed on tall buildings has been utilized in previous research studies to assess the impacts of wind on structures. The wind tunnel test is a primary research method commonly used to quantify the aerodynamic characteristics of high-rise buildings. Wind pressure is measured by placing pressure sensor taps at different locations on tall buildings, and the collected data are used for analysis. However, sensors may malfunction and produce erroneous data; these data losses make it difficult to analyze aerodynamic properties. Therefore, it is essential to generate missing data relative to the original data obtained from neighboring pressure sensor taps at various intervals. This study proposes a deep learning-based, deep convolutional generative adversarial network (DCGAN) to restore missing data associated with faulty pressure sensors installed on high-rise buildings. The performance of the proposed DCGAN is validated by using a standard imputation model known as the generative adversarial imputation network (GAIN). The average mean-square error (AMSE) and average R-squared (ARSE) are used as performance metrics. The calculated ARSE values by DCGAN on the building model's front, backside, left, and right sides are 0.970, 0.972, 0.984 and 0.978, respectively. The AMSE produced by DCGAN on four sides of the building model is 0.008, 0.010, 0.015 and 0.014. The average standard deviation of the actual measures of the pressure sensors on four sides of the model were 0.1738, 0.1758, 0.2234 and 0.2278. The average standard deviation of the pressure values generated by the proposed DCGAN imputation model was closer to that of the measured actual with values of 0.1736,0.1746,0.2191, and 0.2239 on four sides, respectively. In comparison, the standard deviation of the values predicted by GAIN are 0.1726,0.1735,0.2161, and 0.2209, which is far from actual values. The results demonstrate that DCGAN model fits better for data imputation than the GAIN model with improved accuracy and fewer error rates. Additionally, the DCGAN is utilized to estimate the wind pressure in regions of buildings where no pressure sensor taps are available; the model yielded greater prediction accuracy than GAIN.