• Journal of Korean Electronics
• /
• v.11 no.1
• /
• pp.70-74
• /
• 1991

• Composites Research
• /
• v.22 no.4
• /
• pp.1-12
• /
• 2009

In this paper, quasi-static crushing tests of composite circular tubes under axial compression load are conducted to investigate the energy absorption characteristics. Circular tubes used for this experiment are glass/epoxy (GFRP) composite tubes which are fabricated by the filament winding method. One edge of the composite tube is chamfered to reduce the initial peak load and to prevent catastrophic failure during crushing process. Energy absorption characteristics vary significantly according to the constituent materials, fabrication conditions, tube geometry and test condition. In tube geometry, according as inner diameter increase, unstable crush mode is caused by local buckling of delamination, but control of the fiber orientation should help composite tubes get stable crush mode.

• Korean Chemical Engineering Research
• /
• v.57 no.2
• /
• pp.274-282
• /
• 2019

In recent years, big data analysis has been expanded to include automatic control through reinforcement learning as well as prediction through modeling. Research on the utilization of image data is actively carried out in various industrial fields such as chemical, manufacturing, agriculture, and bio-industry. In this paper, we applied NASNet, which is an AutoML reinforced learning algorithm, to DeepU-Net neural network that modified U-Net to improve image semantic segmentation performance. We used BRATS2015 MRI data for performance verification. Simulation results show that DeepU-Net has more performance than the U-Net neural network. In order to improve the image segmentation performance, remove dropouts that are typically applied to neural networks, when the number of kernels and filters obtained through reinforcement learning in DeepU-Net was selected as a hyperparameter of neural network. The results show that the training accuracy is 0.5% and the verification accuracy is 0.3% better than DeepU-Net. The results of this study can be applied to various fields such as MRI brain imaging diagnosis, thermal imaging camera abnormality diagnosis, Nondestructive inspection diagnosis, chemical leakage monitoring, and monitoring forest fire through CCTV.

• Journal of the Economic Geographical Society of Korea
• /
• v.1 no.2
• /
• pp.5-19
• /
• 1998

This attempts to study the change of land use pattern and to (md out the factors to impact the change of the pattern in metropolitan area of Seoul. The data are collected from the 9 units of geomorphological map of the study area with the help of Mapinfo techniques. The data are analyzed statistically with aids of SAS programs. Land use patterns are classified into two: rural and urban and population, urbanization, transportation, industrialization and land development programs are selected as independent variables to change the land use patterns from 1960-1990. The results may be summarized as follows : (1) Arable lands consisted of 30% of the total land in 1960 but the ratio of the arable land decreased to less than 25% in 1990 in the study area. (2) Urban land use types are dominant around southern part of Seoul but rural one are dominant around northern and eastern area of Seoul. (3) Rural type are influenced by population factor but urban land use type are related to transportation and population factors. Land development program is not a significant one to impact the land use pattern in the study area.

• Tunnel and Underground Space
• /
• v.33 no.1
• /
• pp.10-28
• /
• 2023

Long-term shear behavior of the rock discontinuities should be analyzed and its stability should be evaluated to ensure the long-term stability of a high-level radioactive waste disposal repository. The long-term shear behavior of the discontinuities can be modeled with creep and RSF models. The shear creep test, velocity step test, and slide-hold-slide test can be performed to determine their model parameters or analyze the shear behavior by experiments under various conditions. Testing apparatuses for direct shear, triaxial compression, and biaxial shear were mainly used and improved to reproduce the thermo-hydro-mechanical conditions of local bedrock, and it was confirmed that the shear behavior could vary. In order to design a high-level radioactive waste disposal site in Korea, the long-term behavior of rock discontinuities should be investigated in consideration of rock types, thermo-hydro-mechanical conditions, metamorphism, and restoration of shear resistance.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Journal of the Korean Institute of Gas
• /
• v.23 no.4
• /
• pp.10-18
• /
• 2019

The Interior finishing materials tried to evaluate the combustion characteristics and the dangerous characteristic of Floor finish and Wall finish. We often use, conducting the experiment ISO 5660-1, Cone Calorimeter method, and ISO-TR-9122 FT-IR. According to the result of Cone calorimeter experiment, the tile carpet FF3 of Floor material had the highest THR $74.6mj/m^2$ because of the highest risk, and the PHRR of FF1 was $726kW/m^2$, which was easy to bum. As a result FT-IR test, The CO, $CO_2$ ratio was 8,146 PPM for roll carpet FF1 than tile carpet FF2, FF3 5,996, 5,171 PPM, which was a carpet with a high toxicity risk. In the case of wall finishes, The MDF plate(WF3) was THR $86.7mj/m^2$ with a high risk, PHRR $384kW/m^2$ was easy to ignite and toxicity index was 5.5. The CO, $CO_2$ ratio was 1,340~8,596 PPM, But the WF4 was the most toxic with 8,596 PPM.