• Journal of the Korean Society of Safety
• /
• v.37 no.2
• /
• pp.18-27
• /
• 2022

Construction accidents are difficult to prevent because several different types of activities occur simultaneously. The current method of accident analysis only indicates the number of occurrences for one or two variables and accidents have not reduced as a result of safety measures that focus solely on individual variables. Even if accident data is analyzed to establish appropriate safety measures, it is difficult to derive significant results due to a large number of data variables, elements, and qualitative records. In this study, in order to simplify the analysis and approach this complex problem logically, data preprocessing techniques, such as latent class cluster analysis (LCCA) and predictor importance were used to discover the most influential variables. Finally, the correlation was analyzed using an alluvial flow diagram consisting of seven variables and fourteen elements based on accident data. The alluvial diagram analysis using reduced variables and elements enabled the identification of accident trends into four categories. The findings of this study demonstrate that complex and diverse construction accident data can yield relevant analysis results, assisting in the prevention of accidents.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.37 no.3
• /
• pp.304-308
• /
• 2024

In this paper, in order to analyze the PMU data of the accident section, we collected the raw data of a total of 35 PMU installed at the Yeonggwang substation and tried to find a way to analyze the data, and analyzed the data using Excel format and formula. As a result, the three-phase voltage and current data of the PMU were calculated using formulas in Excel and interpreted as effective and reactive power, and it was possible to check the effective and reactive power of the accident section through the graph to see why it was different from before the accident. As a result, it was confirmed that each power was greatly reduced in the graph of the effective and reactive power of the accident section, and it was confirmed that the loss occurred as the power of the accident section was greatly reduced.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.209-212
• /
• 2005

Hyperspectral data consists of more than 200 spectral bands that are highly correlated. In order to utilize hyperspectral data for classification, dimensional reduction or feature extraction is desired. By applying feature extraction, computational complexity of classification can be reduced and classification accuracy may be improved. In this paper, a genetic algorithm based feature extraction technique is proposed. Measure from discriminant analysis is utilized as optimization criterion. A subset of spectral bands is selected by genetic algorithm. Dimension of feature space is further reduced by linear transformation. Feasibility of the proposed technique is evaluated with AVIRIS data.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2008.10a
• /
• pp.352-354
• /
• 2008

A 15.4" WXGA TFT-LCD, featuring integrated a-Si:H gate driver circuits and reduced data driver ICs, has been developed. To reduce number of data lines into 1/2 of conventional structure, the pixel array has been re-mapped with re-organized data signal. Unintended artificial effects such as flicker were removed by adopting the novel pixel array having a 'zigzag' map. To minimize the power consumption, a column inversion method was incorporated in the zigzag pixel array (Fig.1) without modifying the polarity map of conventional dot inversion method.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2000.05a
• /
• pp.842-845
• /
• 2000

It is essential to reduce stiction between the slider and the disk for super-smooth media such as glass disk for high density recording. We developed a stiction-reduced slider by fabricating mechanical bumps on the air bearing surface of the slider by indentation technique. This paper presents a possibility and concept of Stiction-Free-Slider which can operate on the data zone of a magnetic disk. The slider hat many bumps and their heights are in the tens of nm range. The SFS shows good performance on the data zone. Moreover, little wear of the bumps was observed when the preload was small.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.31 no.2
• /
• pp.69-85
• /
• 2006

Transductive Support Vector Machine(TSVM) is one of semi-supervised learning algorithms which exploit the domain structure of the whole data by considering labeled and unlabeled data together. Although it was proposed several years ago, there has been no efficient algorithm which can handle problems with more than hundreds of training examples. In this paper, we propose an efficient branch-and-bound algorithm which can solve large-scale TSVM problems with thousands of training examples. The proposed algorithm uses two bounding techniques: min-cut bound and reduced SVM bound. The min-cut bound is derived from a capacitated graph whose cuts represent a lower bound to the optimal objective function value of the dual problem. The reduced SVM bound is obtained by constructing the SVM problem with only labeled data. Experimental results show that the accuracy rate of TSVM can be significantly improved by learning from the optimal solution of TSVM, rather than an approximated solution.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.2
• /
• pp.929-944
• /
• 2019

Many critical applications require accurate real-time human action recognition. However, there are many hurdles associated with capturing and pre-processing image data, calculating features, and classification because they consume significant resources for both storage and computation. To circumvent these hurdles, this paper presents a recognition machine learning (ML) based system model which uses reduced data structure features by projecting real 3D skeleton modality on virtual 2D space. The MMU VAAC dataset is used to test the proposed ML model. The results show a high accuracy rate of 97.88% which is only slightly lower than the accuracy when using the original 3D modality-based features but with a 75% reduction ratio from using RGB modality. These results motivate implementing the proposed recognition model on an embedded system platform in the future.

• Journal of Biosystems Engineering
• /
• v.18 no.3
• /
• pp.239-251
• /
• 1993

Data acquisition system and computer program developed in this study could be well used in engine vibration analysis. The system and program developed were also operated to be able to control measuring interval, number of channels, number of data. The flywheel was specially studied to provide the proper weight with rubber damper for the engine design at low level of vibration. This study was conducted to obtain basic data which affect the engine vibration. The experiment of this study was performed on original weight flywheel, weight-reduced flywheel, weight-reduced and rubber-coated flywheel, weight-reduced and damper-attached flywheel. Avarage of peak value, maximum vibration, power spectrum density based on FFT analysis are major factors of this experiment. Results were obtained as follows : 1. When rubber was inserted in the flywheel rim of which weight was reduced from 32.2kgf to 24.4 kgf, maximum vibration of the engine was decreased 48.3% at X axis, 35.5% at Y axis and 34.6% at Z axis in comparison with the flywheel of original weight. 2. When the flywheel of rubber damper was compared with the original flywheel, the average of absolute vibration for rubber damped flywheel was decreased at X, Y, Z axis and especially its decreasing rate was so high at X-axis comparing with the other flywheel, which implied that rubber damper was very useful to reducing the vibration of the engine at X axis. 3. Hysteresis losses of X, Y, Z axis were greatly decreased in the flywheel with rubber damper on rim. 4. Damped oscillation effect on X and Y axis vibration above average peak vibration by the flywheel of rubber damper on rim was larger than those by the other flywheels. 5. Power spectrums of vibration at real and imaginery part were bi-mode type. The vibration frequency of rubber dampered flywheel which weight is decreased was slightly increased as compared with original flywheel.

• Smart Structures and Systems
• /
• v.10 no.4_5
• /
• pp.411-426
• /
• 2012

The Canton Tower (formerly named Guangzhou New TV Tower) of 610 m high has been instrumented with a long-term structural health monitoring (SHM) system consisting of over 700 sensors of sixteen types. Under the auspices of the Asian-Pacific Network of Centers for Research in Smart Structures Technology (ANCRiSST), an SHM benchmark problem for high-rise structures has been developed by taking the instrumented Canton Tower as a host structure. This benchmark problem aims to provide an international platform for direct comparison of various SHM-related methodologies and algorithms with the use of real-world monitoring data from a large-scale structure, and to narrow the gap that currently exists between the research and the practice of SHM. This paper first briefs the SHM system deployed on the Canton Tower, and the development of an elaborate three-dimensional (3D) full-scale finite element model (FEM) and the validation of the model using the measured modal data of the structure. In succession comes the formulation of an equivalent reduced-order FEM which is developed specifically for the benchmark study. The reduced-order FEM, which comprises 37 beam elements and a total of 185 degrees-of-freedom (DOFs), has been elaborately tuned to coincide well with the full-scale FEM in terms of both modal frequencies and mode shapes. The field measurement data (including those obtained from 20 accelerometers, one anemometer and one temperature sensor) from the Canton Tower, which are available for the benchmark study, are subsequently presented together with a description of the sensor deployment locations and the sensor specifications.

• Journal of the Society of Naval Architects of Korea
• /
• v.61 no.1
• /
• pp.8-18
• /
• 2024

The present study concerns reduced order modeling of a marine diesel engine, which can be used for outlier detection in status monitoring and carbon intensity index calculation. Principal Component Analysis (PCA) is introduced for the reduced order modeling, focusing on the feasibility of detecting and treating nonlinear variables. By cross-correlation, it is found that there are seven non-linear data channels among 23 data channels, i.e., fuel mode, exhaust gas temperature after the turbocharger, and cylinder coolant temperatures. The dataset is handled so that the mean is located at the nominal continuous rating. Polynomial presentation of the dataset is also applied to reflect the linearity between the engine speed and other channels. The first principal mode shows strong effects of linearity of the most data channels to show the linearity of the system. The non-linear variables are effectively explained by other modes. second mode concerns the temperature of the cylinder cooling water, which shows small correlation with other variables. The third and fourth modes correlates the fuel mode and turbocharger exhaust gas temperature, which have inferior linearity to other channels. PCA is proven to be applicable to data given in binary type of fuel mode selection, as well as numerical type data.