• Journal of Energy Engineering
• /
• v.26 no.4
• /
• pp.57-66
• /
• 2017

Since the heat is not as fast as the electric power and the loss is relatively large compared to the electric power, it is not realistic to operate the thermal transfer system with on operation center like electric power trading. In the case of the Korea District Heating Corporation, where all the thermal transfer are currently being made, only two or four adjacent heat-generating power plants are being the heat trading. Therefore, In this paper, we concluded that it is appropriate to divide the integrated operation center for heat trading into several regions, to operate the hub integrated operation power plant in each region to reflect the characteristics of the heat medium and proposed the thermal transfer mechanism among integrated energy operators. Then, we have developed an algorithm that can optimize the heat transaction for the proposed mechanism and applied it to the actual operators to verify the usefulness of the proposed algorithm.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.6
• /
• pp.415-421
• /
• 2023

In this study, the piezoresistive properties of cementitious composites enhanced with carbon nanotubes for improved electrical conductivity were analyzed using a deep learning-based transformer algorithm. Experimental execution was performed in parallel for acquisition of training data. Previous studies on mixture design, specimen fabrication, chemical composition analysis, and piezoresistive performance testing are also reviewed in this paper. Notably, specimens in which fly ash substituted 50% of the binder material were fabricated and evaluated in this study, in addition to carbon nanotube-infused specimens, thereby exploring the potential enhancement of piezoresistive characteristics in conductive cementitious materials. The experimental results showed more stable piezoresistive responses in specimens with fly-ash substituted binder. The transformer model was trained using 80% of the gathered data, with the remaining 20% employed for validation. The analytical outcomes were generally consistent with empirical measurements, yielding an average absolute error and root mean square error between 0.069 to 0.074 and 0.124 to 0.132, respectively.

• Journal of KIISE:Software and Applications
• /
• v.29 no.12
• /
• pp.966-978
• /
• 2002

There are two important problems in improving text classification systems based on machine learning approach. The first one, called "selection problem", is how to select a minimum number of informative documents from a given document collection. The second one, called "composition problem", is how to reorganize selected training documents so that they can fit an adopted learning method. The former problem is addressed in "active learning" algorithms, and the latter is discussed in "boosting" algorithms. This paper proposes a new learning method, called AdaBUS, which proactively solves the above problems in the context of Naive Bayes classification systems. The proposed method constructs more accurate classification hypothesis by increasing the valiance in "weak" hypotheses that determine the final classification hypothesis. Consequently, the proposed algorithm yields perturbation effect makes the boosting algorithm work properly. Through the empirical experiment using the Routers-21578 document collection, we show that the AdaBUS algorithm more significantly improves the Naive Bayes-based classification system than other conventional learning methodson system than other conventional learning methods

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.4
• /
• pp.71-81
• /
• 2010

In grid system, Task scheduling based on list scheduling models has showed low complexity and high efficiency in fully connected processor set environment. However, earlier schemes did not consider sufficiently the communication cost among tasks and the composition process of lightpath for communication in optical gird environment. In this thesis, we propose LSOG (Leveling Selection in Optical Grid) which sets task priority after forming a hierarchical directed acyclic graph (DAG) that is optimized in optical grid environment. To determine priorities of task assignment in the same level, proposed algorithm executes the task with biggest communication cost between itself and its predecessor. Then, it considers the shortest route for communication between tasks. This process improves communication cost in scheduling process through optimizing link resource usage in optical grid environment. We compared LSOG algorithm with conventional ELSA (Extended List Scheduling Algorithm) and SCP (Scheduled Critical Path) algorithm. We could see the enhancement in overall scheduling performance through increment in CCR value and smoothing network environment.

• Journal of IKEEE
• /
• v.19 no.2
• /
• pp.193-200
• /
• 2015

In this paper, we propose the radiation measuring algorithm and the device composition using CMOS image sensor. The radiation measuring algorithm using CMOS image sensor is based on the radiation particle distinguishing algorithm projected to the CMOS image sensor and accumulated and average number of pixels of the radiation particles projected to dozens of images per second with CMOS image sensor. The radiation particle distinguishing algorithm projected to the CMOS image sensor measures the radiation particle images by dividing them into R, G and B and adjusting the threshold value that distinguishes light intensity and background from the particle of each image. The radiation measuring algorithm measures radiation with accumulated and average number of radiation particles projected to dozens of images per second with CMOS image sensor according to the preset cycle. The hardware devices to verify the suggested algorithm consists of CMOS image sensor and image signal processor part, control part, power circuit part and display part. The test result of radiation measurement using the suggested CMOS image sensor is as follows. First, using the low-cost CMOS image sensor to measure radiation particles generated similar characteristics to that from measurement with expensive GM Tube. Second, using the low-cost CMOS image sensor to measure radiation presented largely similar characteristics to the linear characteristics of expensive GM Tube.

• Journal of Preventive Medicine and Public Health
• /
• v.43 no.6
• /
• pp.479-485
• /
• 2010

Objectives: The Korean Genome and Epidemiology Study (KoGES), a multicenter-based multi-cohort study, has collected information on body composition using two different bioelectrical impedence analysis (BIA) machines. The aim of the study was to evaluate the possibility of whether the test values measured from different BIA machines can be integrated through statistical adjustment algorithm under excellent inter-rater reliability. Methods: We selected two centers to measure inter-rater reliability of the two BIA machines. We set up the two machines side by side and measured subjects' body compositions between October and December 2007. Duplicated test values of 848 subjects were collected. Pearson and intra-class correlation coefficients for inter-rater reliability were estimated using results from the two machines. To detect the feasibility for data integration, we constructed statistical compensation models using linear regression models with residual analysis and R-square values. Results: All correlation coefficients indicated excellent reliability except mineral mass. However, models using only duplicated body composition values for data integration were not feasible due to relatively low $R^2$ values of 0.8 for mineral mass and target weight. To integrate body composition data, models adjusted for four empirical variables that were age, sex, weight and height were most ideal (all $R^2$ > 0.9). Conclusions: The test values measured with the two BIA machines in the KoGES have excellent reliability for the nine body composition values. Based on reliability, values can be integrated through algorithmic statistical adjustment using regression equations that includes age, sex, weight, and height.

• Korean Journal of Mineralogy and Petrology
• /
• v.37 no.2
• /
• pp.23-34
• /
• 2024

It is essential to know the mineral composition of core samples to assess the possibility of gas hydrate (GH) in sediments. During the exploration of gas hydrates (GH), mineral composition values were obtained from each core sample collected in the Ulleung Basin using X-ray diffraction (XRD). Based on this data, machine learning was performed with 3100 input values representing XRD peak intensities and 12 output values representing mineral compositions. The 488 data points were divided into 307 training samples, 132 validation samples, and 49 test samples. The random forest (RF) algorithm was utilized to obtain results. The machine learning results, compared with expert-predicted mineral compositions, revealed a Mean Absolute Error (MAE) of 1.35%. To enhance the performance of the developed model, principal component analysis (PCA) was employed to extract the key features of XRD peaks, reducing the dimensionality of input data. Subsequent machine learning with the refined data resulted in a decreased MAE, reaching a maximum of 1.23%. Additionally, the efficiency of the learning process improved over time, as confirmed from a temporal perspective.

• Proceedings of the KIEE Conference
• /
• 2002.07b
• /
• pp.1229-1231
• /
• 2002

This paper presents the design of DC protection system of Light Rail Transit system. Especially, the composition and interface for DC Switchgear. Digital protection unit and sort of protection algorithm are focused. DC Switchgear (DCSWGR) for LRT testline consist of 5 different panels with peculiar characteristics are examined. Also Basic actuation principle for DC fault select relay (50F), Line Test Device (LTD), DC Overcurrent(OCR) relay are introduced.

• The Journal of Information Technology and Database
• /
• v.2 no.2
• /
• pp.71-85
• /
• 1995

Many research results conducted by neural network researchers have claimed that the classification accuracy of neural networks is superior to, or at least equal to that of conventional methods. However, in series of neural network classifications, it was found that the classification accuracy strongly depends on the characteristics of training data set. Even though there are many research reports that the classification accuracy of neural networks can be different, depending on the composition and architecture of the networks, training algorithm, and test data set, very few research addressed the problem of classification accuracy when the basic assumption of data monotonicity is violated, In this research, development project of automated credit evaluation system is described. The finding was that arrangement of training data is critical to successful implementation of neural training to maintain monotonicity of the data set, for enhancing classification accuracy of neural networks.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.231-233
• /
• 2004

Queuing is waiting lines which play routing service when packet entered. Queuing is decide how and whom is going to provide priority service. This is kind of first in first out(FIFO) or weighted fair queuing(WFQ) method. In this study, UDP design using WFQ way to serve to provide service evenly and rapidly in network. Also in actuality internet, datagram analyzed by packet captured. Queuing services through the requesting port number, input, output, output queuing creation & delete, message request by internet control message protocol(ICMP). Queuing designed in control block module, input queues, input/output module composition. In conclusion, I have confirm queuing result of WFQ method by the datagram information analyzed.