• Smart Structures and Systems
• /
• v.23 no.5
• /
• pp.507-520
• /
• 2019

In the structural health monitoring field, damage detection has been commonly carried out based on the structural model and the engineering features related to the model. However, the extracted features are often subjected to various errors, which makes the pattern recognition for damage detection still challenging. In this study, an automated damage identification method is presented for hanger cables in a tied-arch bridge using a convolutional neural network (CNN). Raw measurement data for Fourier amplitude spectra (FAS) of acceleration responses are used without a complex data pre-processing for modal identification. A CNN is a kind of deep neural network that typically consists of convolution, pooling, and fully-connected layers. A numerical simulation study was performed for multiple damage detection in the hangers using ambient wind vibration data on the bridge deck. The results show that the current CNN using FAS data performs better under various damage states than the CNN using time-history data and the traditional neural network using FAS. Robustness of the present CNN has been proven under various observational noise levels and wind speeds.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.4
• /
• pp.482-488
• /
• 2013

Load forecasting is essential to the economical and the stable power system operations. In general, the forecasting days can be classified into weekdays, weekends, special days and special light-load periods in short-term load forecast. Special light-load periods are the consecutive holidays such as Lunar New Years holidays, Korean Thanksgiving holidays and summer special light-load period. For the weekdays and the weekends forecast, the conventional methods based on the statistics are mainly used and show excellent results for the most part. The forecast algorithms for special days yield good results also but its forecast error is relatively high than the results of the weekdays and the weekends forecast methods. For summer special light-load period, none of the previous studies have been performed ever before so if the conventional methods are applied to this period, forecasting errors of the conventional methods are considerably high. Therefore, short-term load forecast for summer special light-load period have mainly relied on the experience of power system operation experts. In this study, the trends of load profiles during summer special light-load period are classified into three patterns and new forecast algorithms for each pattern are suggested. The proposed method was tested with the last ten years' summer special light-load periods. The simulation results show the excellent average forecast error near 2%.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.3
• /
• pp.202-208
• /
• 2013

In the image processing for VS (Video Surveillance), the detection of moving entities in a monitored scene is an important step. A background subtraction technique has been widely employed to find the moving entities. However, the extracted foreground regions often include not only real entities but also their cast shadows, and this can cause errors in following image processing steps, such as tracking, recognition, and analysis. In this paper, a novel technique is proposed to determine the shadow pixels of moving objects in the foreground image of a VS camera. Compared to existing techniques where the same decision criteria are applied to all moving pixels, the proposed technique determines shadow pixels using local features based on two facts: First, the amount of pixel intensity drop due to a shadow depends on the intensity level of background. Second, the distribution pattern of pixel intensities remains even if a shadow is cast. The proposed method has been tested at various situations with different backgrounds and moving humans in different colors.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.11
• /
• pp.1455-1462
• /
• 2018

As the interests in eco-friendly energy has increased, the interests in Electric Vehicles(EVs) are increasing as well. Moreover, due to the government's economic support for EVs, penetration level of it has rapidly increased. These sharp increases, however, induce various problems in distribution system, such as voltage/frequency variations, peak demand increasement, demand control, etc. To minimize these possible matters, lots of research have conducted. Nevertheless, most of it assumed extremely important factors, such as numbers and charging patterns of EVs. It inevitably results in errors in their research, and thus make it difficult to prevent the possible matters from EVs. In this paper, therefore, we use actual EVs charging data from KEPCO, and analysis and deduction of it were conducted. The simulations were carried out for four aspect(season, region, purpose).

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.7 no.4
• /
• pp.221-228
• /
• 2007

In this paper, we propose a new symbolic simulation for scan chain diagnosis to solve the diagnosis resolution problem. The proposed scan chain fault simulation, called the SF-simulation, is able to analyze the effects caused by faulty scan cells in good scan chains. A new scan chain fault simulation is performed with a modified logic ATPG pattern. In this simulation, we consider the effect of errors caused by scan shifting in the faulty scan chain. Therefore, for scan chain diagnosis, we use the faulty information in good scan chains which are not contaminated by the faults while unloading scan out responses. The SF-simulation can tighten the size of the candidate list and achieve a high diagnosis resolution by analyzing fault effects of good scan chains, which are ignored by most previous works. Experimental results demonstrate the effectiveness of the proposed method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.12 no.10
• /
• pp.968-974
• /
• 2006

We are moving into the era of ubiquitous computing. Ubiquitous Sensor Network (USN) is a base of such computing paradigm, where recognizing the identification and the position of objects is important. For the object identification, RFID tags are commonly used. For the object positioning, use of sensors such as laser and ultrasonic scanners is popular. Recently, there have been a few attempts to apply RFID technology in robot localization by replacing the sensors with RFID readers to achieve simpler and unified USN settings. However, RFID does not provide enough sensing accuracy for some USN applications such as robot navigation, mainly because of its inaccuracy in distance measurements. In this paper, we describe our approach on achieving accurate navigation using RFID. We solely rely on RFID mechanism for the localization by providing coordinate information through RFID tag installed floors. With the accurate positional information stored in the RFID tag, we complement coordinate errors accumulated during the wheel based robot navigation. We especially focus on how to distribute RFID tags (tag pattern) and how many to place (tag granularity) on the RFID tag-floor. To determine efficient tag granularities and tag patterns, we developed a simulation program. We define the error in navigation and use it to compare the effectiveness of the navigation. We analyze the simulation results to determine the efficient granularities and tag arrangement patterns that can improve the effectiveness of RFID navigation in general.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.68 no.3
• /
• pp.471-479
• /
• 2019

This research proposes a indoor magnetic map matching algorithm that improves the position accuracy by employing multiple magnetic sensors and probabilistic candidate weighting function. Since the magnetic field is easily distorted by the surrounding environment, the distorted magnetic field can be used for position mapping, and multiple sensor configuration is useful to improve mapping accuracy. Nevertheless, the position error is likely to increase because the external magnetic disturbances have repeated pattern in indoor environment and several points have similar magnetic field distortion characteristics. Those errors cause large position error, which reduces the accuracy of the position detection. In order to solve this problem, we propose a method to reduce the error using multiple sensors and likelihood boundaries that uses human walking characteristics. Also, to reduce the maximum position error, we propose an algorithm that weights according to their importance. We performed indoor walking tests to evaluate the performance of the algorithm and analyzed the position detection error rate and maximum distance error. From the results we can confirm that the accuracy of position detection is greatly improved.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.5
• /
• pp.121-127
• /
• 2021

Reverse offset printing is considering as an emerging technology for printed electronics owing to its environmentally friendliness and cost-effectiveness. In reverse offset printing, selecting the materials for cliché and blanket is critical because of its minimum resolution, registration errors, aspect ratio of reliefs, pattern area, and reusability. Various materials such as silicon, quartz, glass, electroplated nickel plates, and imprinted polymers on rigid substrates can be used for the reverse offset printing of cliché. However, when new structures are designed for specific applications, new clichés need to re-fabricated each time employing multiple time-consuming and costly processes. Therefore, by modifying the blanket materials containing the printing ink, several new structures can be easily created using the same cliché. In this study, we investigated various elastomeric materials and evaluated their applicability for designing a highly stretchable blanket with controlled elastic deformation to implement tunable reverse offset printing.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.5
• /
• pp.583-586
• /
• 2004

Transition jitter noise is one of major sources of detection errors in high density recording channels. Implementation complexity of the optimal detector for such channels is high due to the data dependency and correlated nature of the jitter noise. In this paper, two types of hardware efficient sub-optimal detectors are derived by modifying branch metric of Viterbi algorithm and applied to partial response (PR) channels combined with run length limited modulation coding. The additional complexity over the conventional Viterbi algorithm to incorporate the modified branch metric is either a multiplication or an addition for each branch metric in the Viterbi trellis.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2000.06a
• /
• pp.549-555
• /
• 2000

The improved mode superposition methods for non-classically damped systems are presented in this paper. Generally, the mode superposition method uses a relatively small subset of the normal modes of structures. The mode acceleration method and the modal truncation augmentation method improve the results of the mode superposition method by considering effects of truncated high modes. For using these methods to analyze non-classically damped systems, the systems are approximated to the classically damped systems and thereby the errors are induced. In this paper, the mode acceleration method and the modal truncation augmentation method are expanded to analyze the non-classically damped systems. The applicability of the expanded methods is verified by closed form solutions and numerical examples. The expanded modal truncation augmentation method is conditionally stable depending on the pattern of the external loading in the non-classically damped systems whereas the expanded mode acceleration method is stable for the all cases of loading. In the stable case, the results are the same with those of the expanded mode acceleration method.