• Journal of Biosystems Engineering
• /
• v.43 no.3
• /
• pp.229-236
• /
• 2018

Purpose: Makgeolli is a traditional alcoholic beverage made from rice with a fermentation starter called "nuruk." The concentration of alcohol in makgeolli depends on the temperature of the fermentation tank. It is important to monitor the alcohol concentration to manage the makgeolli production process. Methods: Data were collected from 84 makgeolli fermentation tanks over a year period. Independent variables included the temperatures of the tanks and the room where the tanks were located, as well as the quantity, acidity, and water concentration of the source. Software for the multilayer perceptron model (MLP) was written in Python using the Scikit-learn library. Results: Many models were created for which the optimization converged within 100 iterations, and their coefficients of determination $R^2$ were considerably high. The coefficient of determination $R^2$ of the best model with the training set and the test set were 0.94 and 0.93, respectively. The fact that the difference between them was very small indicated that the model was not overfitted. The maximum and minimum error was approximately 2% and the total MSE was 0.078%. Conclusions: The MLP model could help predict the alcohol concentration and to control the production process of makgeolli. In future research, the optimization of the production process will be studied based on the model.

• Journal of the Korean Geotechnical Society
• /
• v.20 no.2
• /
• pp.125-130
• /
• 2004

In this paper, the method of parameter estimation of a mathematical constitutive model blown as the smooth elasto-plastic cap model is studied. To predict the response of the real soil using this model, the eight parameters describing the constitutive equations have to be determined. First, experimental data are obtained from simple laboratory experiments such as one dimensional confined compression test in a consolidometer and drained triaxial compression test with the Ottawa sand f3r the reference value. Then, the numerical experiments are performed in the cap model with initial guessed parameters. The optimization method is utilized to fit the model response to experimental data by minimizing the error between the laboratory and numerical responses. Special attention is given to the parameter estimation procedure of numerical triaxial test due to the difficulty of the lateral strain measurements.

• Korean Journal of Remote Sensing
• /
• v.34 no.1
• /
• pp.141-150
• /
• 2018

Crop type classification is essential for supporting agricultural decisions and resource monitoring. Remote sensing techniques, especially using hyperspectral imagery, have been effective in agricultural applications. Hyperspectral imagery acquires contiguous and narrow spectral bands in a wide range. However, large dimensionality results in unreliable estimates of classifiers and high computational burdens. Therefore, reducing the dimensionality of hyperspectral imagery is necessary. In this study, the Random Forest (RF) classifier was utilized for dimensionality reduction as well as classification purpose. RF is an ensemble-learning algorithm created based on the Classification and Regression Tree (CART), which has gained attention due to its high classification accuracy and fast processing speed. The RF performance for crop classification with airborne hyperspectral imagery was assessed. The study area was the cultivated area in Chogye-myeon, Habcheon-gun, Gyeongsangnam-do, South Korea, where the main crops are garlic, onion, and wheat. Parameter optimization was conducted to maximize the classification accuracy. Then, the dimensionality reduction was conducted based on RF variable importance. The result shows that using the selected bands presents an excellent classification accuracy without using whole datasets. Moreover, a majority of selected bands are concentrated on visible (VIS) region, especially region related to chlorophyll content. Therefore, it can be inferred that the phenological status after the mature stage influences red-edge spectral reflectance.

• Journal of Communications and Networks
• /
• v.15 no.3
• /
• pp.266-275
• /
• 2013

The exact closed-form expressions for outage probability and bit error rate of spectrum sharing-based multi-hop decode-and-forward (DF) relay networks in non-identical Rayleigh fading channels are derived. We also provide the approximate closed-form expression for the system ergodic capacity. Utilizing these tractable analytical formulas, we can study the impact of key network parameters on the performance of cognitive multi-hop relay networks under interference constraints. Using a linear network model, we derive an optimum relay position scheme by numerically solving an optimization problem of balancing average signal-to-noise ratio (SNR) of each hop. The numerical results show that the optimal scheme leads to SNR performance gains of more than 1 dB. All the analytical expressions are verified by Monte-Carlo simulations confirming the advantage of multihop DF relaying networks in cognitive environments.

• Transactions of the Society of Information Storage Systems
• /
• v.3 no.1
• /
• pp.5-12
• /
• 2007

This paper investigates the design of disturbance observer for rotational vibration suppression in disk drive systems. The design aims to provide an optimal controller which satisfies both vibration performance and robust stability. It consists of an inversion method, a special filter, and optimization scheme. Firstly a new inversion method is introduced, which provides more accurate inversion compared to conventional zero phase error method. The inversion is to deal with unstable zeros in the plant model. Secondly a special filter for disturbance selection is given, which features adjustable gain and band pass characteristics so that it enables flexible shaping of the loop considering the trade-off between performance and stability margins. And finally the parameters of disturbance observer are optimized in conjunction with external disturbance model. Simulation and experiment on commercial hard disk drives confirm that the design is very effective to such disturbance which is hard to be handled by conventional approach.

• Journal of Powder Materials
• /
• v.11 no.5
• /
• pp.399-411
• /
• 2004

Crack generation during die compaction and distortion during sintering have been critical problems for the conventional pressing and sintering process. Until now, trial and error approach with engineers' industrial experiences has been only solution to protect the crack generation and distortion. However, with complexity in shape and process it is very difficult to design process conditions without CAE analysis. We developed the exclusive CAE software (PMsolver/Compaction) for die compaction process. The accuracy of PMsolver is verified by comparing the finite element simulation results with experimental results. The simplified procedures to find material properties are proposed and verified with iron based powder and tungsten carbide powder. Based on the accurate simulation result by PMsolver, the optimal process conditions are designed to get uniform density distribution in a powder compact after die compaction process by using a derivative based optimization scheme. In addition, the effect of non-uniform density distribution in a powder compact on distortion during sintering is shown in case of the fabrication of tungsten carbide insert.

• ETRI Journal
• /
• v.32 no.3
• /
• pp.395-405
• /
• 2010

A novel multiclassifier system (MCS) strategy is proposed and applied to a text-dependent speaker verification task. The presented scheme optimizes the linear combination of classifiers on an on-line basis. In contrast to ordinary MCS approaches, neither a priori distributions nor pre-tuned parameters are required. The idea is to improve the most accurate classifier by making use of the incremental information provided by the second classifier. The on-line multiclassifier optimization approach is applicable to any pattern recognition problem. The proposed method needs neither a priori distributions nor pre-estimated weights, and does not make use of any consideration about training/testing matching conditions. Results with Yoho database show that the presented approach can lead to reductions in equal error rate as high as 28%, when compared with the most accurate classifier, and 11% against a standard method for the optimization of linear combination of classifiers.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.6
• /
• pp.411-416
• /
• 2016

This paper presents the design methodology of an unknown input observer for Lipschitz nonlinear systems with unknown inputs in the framework of convex optimization. We use an unknown input observer (UIO) to consider both nonlinearity and disturbance. By deriving a sufficient condition for exponential stability in the linear matrix inequality (LMI) form, existence of a stabilizing observer gain matrix of UIO will be assured by checking whether the quadratic stability margin of the error dynamics is greater than the Lipschitz constant or not. If quadratic stability margin is less than a Lipschitz constant, the coordinate transformation may be used to reduce the Lipschitz constant in the new coordinates. Furthermore, to reduce the maximum singular value of the observer gain matrix elements, an object function to minimize it will be optimally designed by modifying its magnitude so that amplification of sensor measurement noise is minimized via multi-objective optimization algorithm. The performance of UIO is compared to a nonlinear observer (Luenberger-like) with an application to a flexible joint robot system considering a change of load and disturbance. Finally, it is validated via simulations that the estimated angular position and velocity provide true values even in the presence of unknown inputs.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2015.10a
• /
• pp.55-57
• /
• 2015

The tracking system plays a key role in accurate estimation and prediction of maneuvering vessel's position and velocity in a bid to enhance safety by taking avoiding action against collision. Therefore, in order to achieve this, many ocean- going vessels are equipped with radar and the ARPA system. However, the accuracy of prediction highly depends on the choice of the gain parameters, ${\alpha}$, ${\beta}$ and ${\gamma}$ employed in the tracking filter. P revious research of this paper was based on theoretically developing an algorithm for a tracking module. This research paper is hence a continuation by the authors to determine the optimal values of the gain parameters used in the tracking module. A tracking algorithm is developed using the ${\alpha}-{\beta}-{\gamma}$ filter to carry out prediction and smoothing of the positions and velocities. Numerical simulations are then performed to evaluate the optimal values of the smoothing parameters that will improve the performance of the tracking module and reduce measurement noise. The twice distance root mean square (2drms) is then calculated to determine error variation.

• ETRI Journal
• /
• v.33 no.2
• /
• pp.155-162
• /
• 2011

The scalable extension of the H.264/AVC video coding standard (SVC) demonstrates superb adaptability in video communications. Joint source and channel coding (JSCC) has been shown to be very effective for such scalable video consisting of parts of different significance. In this paper, a new JSCC scheme for SVC transmission over packet loss channels is proposed which performs two-dimensional optimization on the quality layers of each frame in a rate-distortion (R-D) sense as well as on the temporal hierarchical structure of frames under dependency constraints. To compute the end-to-end R-D points of a frame, a novel reduced trellis algorithm is developed with a significant reduction of complexity from the existing Viterbi-based algorithm. The R-D points of frames are sorted under the hierarchical dependency constraints and optimal JSCC solution is obtained in terms of the best R-D performance. Experimental results show that our scheme outperforms the existing scheme of [13] with average quality gains of 0.26 dB and 0.22 dB for progressive and non-progressive modes respectively.