• Journal of Positioning, Navigation, and Timing
• /
• v.9 no.4
• /
• pp.397-403
• /
• 2020

In order to synchronize a remote system time to the reference time like Coordinated Universal Time (UTC), it is required to compare the time difference between the two clocks. The time comparison data may have some outliers and the time synchronization performance can be significantly degraded if the outliers are not removed. Therefore, it is required to employ an effective outlier detection algorithm for keeping high accurate system time. In this paper, an outlier detection method is presented for the time difference data of GNSS time transfer receivers. The time difference data between the system time and the GNSS usually have slopes because the remote system clock is under free running until synchronized to the reference clock time. For investigating the outlier detection performance of the proposed algorithm, simulations are performed by using the time difference data of a GNSS time transfer receiver corrected to a free running Cesium clock with intentionally inserted outliers. From the simulation, it is investigated that the proposed algorithm can effectively detect the inserted outliers while conventional methods such as modified Z-score and adjusted boxplot cannot. Furthermore, it is also observed that the synchronization performance can be degraded to more than 15% with 20 outliers compared to that of original data without outliers.

• Journal of Information Processing Systems
• /
• v.8 no.4
• /
• pp.555-566
• /
• 2012

The cloud environment makes it possible to analyze large data sets in a scalable computing infrastructure. In the bioinformatics field, the applications are composed of the complex workflow tasks, which require huge data storage as well as a computing-intensive parallel workload. Many approaches have been introduced in distributed solutions. However, they focus on static resource provisioning with a batch-processing scheme in a local computing farm and data storage. In the case of a large-scale workflow system, it is inevitable and valuable to outsource the entire or a part of their tasks to public clouds for reducing resource costs. The problems, however, occurred at the transfer time for huge dataset as well as there being an unbalanced completion time of different problem sizes. In this paper, we propose an adaptive resource-provisioning scheme that includes run-time data distribution and collection services for hiding the data transfer time. The proposed adaptive resource-provisioning scheme optimizes the allocation ratio of computing elements to the different datasets in order to minimize the total makespan under resource constraints. We conducted the experiments with a well-known sequence alignment algorithm and the results showed that the proposed scheme is efficient for the cloud environment.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.3
• /
• pp.845-850
• /
• 2023

In this paper, we propose a method for diagnosing ball bearing vibration using transfer learning. STFT, which can analyze vibration signals in time-frequency, was used as input to CNN to diagnose failures. In order to rapidly learn CNN-based deep artificial neural networks and improve diagnostic performance, we proposed a transfer learning-based deep learning learning technique. For transfer learning, the feature extractor and classifier were selectively learned using a VGG-based image classification model, the data set for learning was publicly available ball bearing vibration data provided by Case Western Reserve University, and performance was evaluated by comparing the proposed method with the existing CNN model. Experimental results not only prove that transfer learning is useful for condition diagnosis in ball bearing vibration data, but also allow other industries to use transfer learning to improve condition diagnosis.

• Smart Media Journal
• /
• v.9 no.4
• /
• pp.26-35
• /
• 2020

Attempts to predict future stock prices have been studied steadily since the past. However, unlike general time-series data, financial time-series data has various obstacles to making predictions such as non-stationarity, long-term dependence, and non-linearity. In addition, variables of a wide range of data have limitations in the selection by humans, and the model should be able to automatically extract variables well. In this paper, we propose a 'sliding time step normalization' method that can normalize non-stationary data and LSTM autoencoder to compress variables from all variables. and 'moving transfer learning', which divides periods and performs transfer learning. In addition, the experiment shows that the performance is superior when using as many variables as possible through the neural network rather than using only 100 major financial variables and by using 'sliding time step normalization' to normalize the non-stationarity of data in all sections, it is shown to be effective in improving performance. 'moving transfer learning' shows that it is effective in improving the performance in long test intervals by evaluating the performance of the model and performing transfer learning in the test interval for each step.

• Journal of ILASS-Korea
• /
• v.21 no.2
• /
• pp.78-87
• /
• 2016

This study experimentally investigated the effects of droplet temperature on the heat transfer characteristics during collision of a single droplet on a heated wall above the Leidenfrost temperature. Experiments were performed by varying temperature from 40 to $100^{\circ}C$ while the collision velocity and wall temperature were maintained constant at 0.7 m/s at $500^{\circ}C$, respectively. Evolution of temperature distribution at the droplet-wall interface as well as collision dynamics of the droplet were simultaneously recorded using synchronized high-speed video and infrared cameras. The local heat flux distribution at the collision surface was deduced using the measured temperature distribution data. Various physical parameters, including residence time, local heat flux distribution, heat transfer rate, heat transfer effectiveness and vapor film thickness, were measured from the visualization data. The results showed that increase in droplet temperature reduces the residence time and increases the vapor film thickness. This ultimately results in reduction in the total heat transfer by conduction through the vapor film during droplet-wall collision.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.8 no.1
• /
• pp.47-53
• /
• 2012

Sensor networks and car networks that have different structure from that of conventional TCP/IP network require extreme network environment due to frequent change of connectivity. Because such extreme network environment has characteristics like unreliable link connectivity, long delay time, asymmetrical data transfer rate, and high error rate, etc., it is difficult to perform normally with the conventional TCP/P-based routing. DTNs (delay and disruption tolerant network) was designed to support data transfer in extreme network environment with long delay time and no guarantee for continuous connectivity between terminals. This study suggests an algorithm that limits the maximum number of copying transferred message to L by improving the spray and wait routing protocol, which is one of the conventional DTNs routing protocols, and using the azimuth and density data of the mobile nods. The suggested algorithm was examined by using ONE, a DTNs simulator. As a result, it could reduce the delay time and overhead of unnecessary packets compared to the conventional spray and wait routing protocol.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.5
• /
• pp.375-382
• /
• 2016

This paper presents the results for the performance improvement of underwater communication in a passive time reversal mirror (PTRM) using ray-based blind deconvolution (RBD). In conventional PTRM, the signal to be recovered is found from matched-filtering the received probe signal. However, the communication performance is degraded because the time-varying impulse response for each data frame is not reflected in the received probe signal. In this study, the time-variant transfer function is estimated from each received data frame using RBD, and the estimated time-variant transfer function is then used to recover the data signal using PTRM. The results from the experimental data show that the suggested method improves the communication performance when comparing with the conventional PTRM.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2298-2303
• /
• 2003

In this paper, very simple methods based on the exponential analysis are presented by which transfer function models for processes can easily be obtained. These methods employ step responses or impulse responses of the processes. These can also give a more precise transfer function model compared to the well-known graphical methods. Transfer functions are determined based on Prony method, which is one of the oldest and the most representative methods in the exponential analysis. Here, the method is reformed and applied to obtain the so-called low-order transfer function with pure time delay from the data of the step response. The effectiveness of the proposed method is examined through several numerical examples and experiments of the 2-tank level control process.

• Korean Journal of Child Studies
• /
• v.7 no.1
• /
• pp.67-84
• /
• 1986

The purpose of the present research was to investigate cross-modal transfer, especially tactual-to-visual transfer in infancy and to study the relation between failure of cross-modal transfer performance and length of familiarization period. The subjects of this study were 60 infants, 10 boys and 10 girls at each level: six, nine, and twelve months of age. All were normal, healthy, full-term babies. The mothers' educational achievement was controlled at more than 12 years of schooling. There were two separate experimental conditions, one 30-sec and one 60-sec familiarization period. Each experimental condition consisted of a tactual familiarization and a visual recognition memory test. Each child was presented with these 2 sets of cross-modal stimuli in one of the 2 experimental conditions. Infants' visual responses in the visual recognition memory test were videotaped for 20 seconds. Visual fixation time to novel and familiar stimuli was observed throughout the test. The data was analyzed with t-test, percentage of total fixation time to novel stimuli, and ANOVA. The results showed that: 1) Significant differences were found in the cross-modal transfer performance from touch to vision between the 3 age groups. This is, 6 and 9 month old infants didn't show cross-modal transfer in the 30-sec condition, but 12 month old infants did show cross-modal transfer in the 30-second. 2) In all 3 age groups, no significant differences were found in cross-modal transfer performance between the two conditions.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.7 no.4
• /
• pp.461-468
• /
• 1983

The series of experiments have been performed to study the transient heat transfer in air from a suddenly heated vertical thin wire. A platinum wire has been used as a resistance thermometer as well as a heating element to eliminate the disturbances in the measurements. The measured temperature as a function of time is compared with the calculated transient temperature with the aid of a pure conduction equation. The overshoot phenomena in terms of the Nusselt numbers have been detected and it is reasonable to define the delay time at which the onset of convection heat transfer occurs. The measured data are compared with the existing steady-state data and the agreements are reasonable within the comparable ranges.