• Journal of the Korea institute for structural maintenance and inspection
• /
• v.20 no.4
• /
• pp.1-8
• /
• 2016

Recently, buildings tend to be large size, complex shape and functional. As the size of buildings is becoming massive, the need for structural health monitoring(SHM) technique is ever-increasing. Various SHM techniques have been studied for buildings which have different dynamic characteristics and are influenced by various external loads. Generally, the visual inspection and non-destructive test for an accessible point of structures are performed by experts. But nowadays, the system is required which is online measurement and detect risk elements automatically without blind spots on structures. In this study, in order to consider the response of non-linear structures, proposed a signal feature extraction and the adaptive threshold setting algorithm utilized to determine the abnormal behavior by using statistical methods such as control chart, root mean square deviation, generalized extremely distribution. And the performance of that was validated by using the acceleration response of structures during earthquakes measuring system of forced vibration tests and actual operation.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.10a
• /
• pp.1108-1111
• /
• 2009

Auto Pilot is the system which move automatically the vessel through locating operation mode to automatic after entering operating course using a electronic chart or plotter. And water jet is the a propulsion system that make a power to push the vessel through spouting the accelerated water which is absorbed by the hole in the bottom of vessel. The water jet receive the effect of the depth of water lowly, it's acceleration efficiency is higher under high speed and have an advantage on vibrating and floating sound, so it's demand is increasing as new propulsion system. However, the signal systems of auto pilot and water jet are different, we need the system to interface between each system. We designed the interface that efficiently digital feed back control embedded module between auto pilot and water jet system in this paper.

• Smart Structures and Systems
• /
• v.23 no.3
• /
• pp.307-318
• /
• 2019

The frequently excessive vibrations presented in civil structures during seismic events or service conditions may result in users' discomfort, or worst, in structures failure, producing economic and even human casualties. This work contributes in proposing the synthesis of a nonlinear optimal control strategy for semiactive structural control, with the main characteristic that the synthesis considers both the structure model and the semiactive actuator nonlinear dynamics, which produces a nonlinear system that requires a nonlinear controller design. The aim is to reduce the unwanted vibrations in the response of civil structures, by means of intelligent fluid semiactive actuator such as the Magnetorheological Damper (MRD), which is a device with a low level of power consumption. The civil structures for which the proposed control methodology can be applied are those admitting a state-dependent coefficient factorized representation model, such as buildings, bridges, among others. A scaled model of a three storey building is analyzed as a case study, whose dynamical response involves displacement, velocity and acceleration of each one of the storeys, subjected to the North-South component of the September 19th., 2017, Puebla-Morelos (7.1M), Mexico earthquake. The investigation rests on comparing the structural response over time for two different conditions: with no control device installed and with one MRD installed between the first floor and the ground, where a nonlinear optimal signal for the MRD input voltage is determined. Simulation results are presented to show the effectiveness of the proposed controller for reducing the building's dynamical response.

• Journal of Sensor Science and Technology
• /
• v.28 no.5
• /
• pp.311-317
• /
• 2019

Rotating devices are commonly installed in power plants and factories. This study proposes a self-powered sensor node that is powered by converting the vibration energy of a rotating device into electrical energy. The self-powered sensor consists of a piezoelectric harvester for self-power generation, a rectifier circuit to rectify the AC signal, a sensor unit for measuring the vibration frequency, and a circuit to control the light emitting diode (LED) lighting. The frequency of the vibration source was measured using a piezoelectric-cantilever-type vibration frequency sensor. A green LED was illuminated when the measured frequency was within the normal range. The power generated by the piezoelectric harvester was determined, and the LED operation was assessed in terms of the vibration frequency. The piezoelectric harvester was found to generate a power of 3.061 mW or greater at a vibration acceleration of 1.2 g ($1g=9.8m/s^2$) and vibration frequencies between 117 and 123 Hz. Notably, the power generated was 4.099 mW at 122 Hz. As such, our self-powered sensor node can be used as a module for monitoring rotating devices, because it can convert vibration energy into electrical energy when installed on rotating devices such as air compressors.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.9
• /
• pp.1224-1230
• /
• 2020

Since the spread of smart phones, interest in wearable devices has increased and diversified, and is closely related to the lives of users, and has been used as a method for providing personalized services. In this paper, we propose a method to detect the user's behavior by applying information from a 3-axis acceleration sensor and a 3-axis gyro sensor embedded in a smartphone to a convolutional neural network. Human behavior differs according to the size and range of motion, starting and ending time, including the duration of the signal data constituting the motion. Therefore, there is a performance problem for accuracy when applied to a convolutional neural network as it is. Therefore, we proposed a Time-Division Feature Fusion Convolutional Neural Network (TDFFCNN) that learns the characteristics of the sensor data segmented over time. The proposed method outperformed other classifiers such as SVM, IBk, convolutional neural network, and long-term memory circulatory neural network.

• Journal of Advanced Navigation Technology
• /
• v.25 no.4
• /
• pp.299-304
• /
• 2021

Existing human activity recognition systems detect activities through devices such as wearable sensors and cameras. However, these methods require additional devices and costs, especially for cameras, which cause privacy issue. Using WiFi signals that are already installed can solve this problem. In this paper, we propose a CNN-based human activity recognition system using channel state information of WiFi signals, and present results of designing and implementing accelerated hardware structures. The system defined four possible behaviors during studying in indoor environments, and classified the channel state information of WiFi using convolutional neural network (CNN), showing and average accuracy of 91.86%. In addition, for acceleration, we present the results of an accelerated hardware structure design for fully connected layer with the highest computation volume on CNN classifiers. As a result of performance evaluation on FPGA device, it showed 4.28 times faster calculation time than software-based system.

• Journal of Microbiology and Biotechnology
• /
• v.32 no.10
• /
• pp.1245-1252
• /
• 2022

Induced pluripotent stem cells (iPSCs) can be generated from somatic cells using Oct4, Sox2, Klf4, and c-Myc (OSKM). Small molecules can enhance reprogramming. Licochalcone D (LCD), a flavonoid compound present mainly in the roots of Glycyrrhiza inflata, acts on known signaling pathways involved in transcriptional activity and signal transduction, including the PGC1-α and MAPK families. In this study, we demonstrated that LCD improved reprogramming efficiency. LCD-treated iPSCs (LCD-iPSCs) expressed pluripotency-related genes Oct4, Sox2, Nanog, and Prdm14. Moreover, LCD-iPSCs differentiated into all three germ layers in vitro and formed chimeras. The mesenchymal-to-epithelial transition (MET) is critical for somatic cell reprogramming. We found that the expression levels of mesenchymal genes (Snail2 and Twist) decreased and those of epithelial genes (DSP, Cldn3, Crb3, and Ocln) dramatically increased in OR-MEF (OG2^+/+/ROSA26^+/+) cells treated with LCD for 3 days, indicating that MET effectively occurred in LCD-treated OR-MEF cells. Thus, LCD enhanced the generation of iPSCs from somatic cells by promoting MET at the early stages of reprogramming.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.181-193
• /
• 2022

Data-driven structural health monitoring (SHM) of civil infrastructure can be used to continuously assess the state of a structure, allowing preemptive safety measures to be carried out. Long-term monitoring of large-scale civil infrastructure often involves data-collection using a network of numerous sensors of various types. Malfunctioning sensors in the network are common, which can disrupt the condition assessment and even lead to false-negative indications of damage. The overwhelming size of the data collected renders manual approaches to ensure data quality intractable. The task of detecting and classifying an anomaly in the raw data is non-trivial. We propose an approach to automate this task, improving upon the previously developed technique of image-based pre-processing on one-dimensional (1D) data by enriching the features of the neural network input data with multiple channels. In particular, feature engineering is employed to convert the measured time histories into a 3-channel image comprised of (i) the time history, (ii) the spectrogram, and (iii) the probability density function representation of the signal. To demonstrate this approach, a CNN model is designed and trained on a dataset consisting of acceleration records of sensors installed on a long-span bridge, with the goal of fault detection and classification. The effect of imbalance in anomaly patterns observed is studied to better account for unseen test cases. The proposed framework achieves high overall accuracy and recall even when tested on an unseen dataset that is much larger than the samples used for training, offering a viable solution for implementation on full-scale structures where limited labeled-training data is available.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.77-91
• /
• 2022

Various monitoring systems have been implemented in civil infrastructure to ensure structural safety and integrity. In long-term monitoring, these systems generate a large amount of data, where anomalies are not unusual and can pose unique challenges for structural health monitoring applications, such as system identification and damage detection. Therefore, developing efficient techniques is quite essential to recognize the anomalies in monitoring data. In this study, several machine learning techniques are explored and implemented to detect and classify various types of data anomalies. A field dataset, which consists of one month long acceleration data obtained from a long-span cable-stayed bridge in China, is employed to examine the machine learning techniques for automated data anomaly detection. These techniques include the statistic-based pattern recognition network, spectrogram-based convolutional neural network, image-based time history convolutional neural network, image-based time-frequency hybrid convolution neural network (GoogLeNet), and proposed ensemble neural network model. The ensemble model deliberately combines different machine learning models to enhance anomaly classification performance. The results show that all these techniques can successfully detect and classify six types of data anomalies (i.e., missing, minor, outlier, square, trend, drift). Moreover, both image-based time history convolutional neural network and GoogLeNet are further investigated for the capability of autonomous online anomaly classification and found to effectively classify anomalies with decent performance. As seen in comparison with accuracy, the proposed ensemble neural network model outperforms the other three machine learning techniques. This study also evaluates the proposed ensemble neural network model to a blind test dataset. As found in the results, this ensemble model is effective for data anomaly detection and applicable for the signal characteristics changing over time.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.3
• /
• pp.73-82
• /
• 2022

Driver acceptance of autonomous driving is very important. The autonomous driving longitudinal controller, which is one of the factors affecting acceptability, consists of a high-level controller and a low-level controller. The host controller decides the cruise control and the space control according to the situation and creates the required target speed. The sub-controller performs control by creating an acceleration signal to follow the target speed. In this paper, we propose an algorithm to improve the inter-vehicle distance fluctuations that occur in the cruise control and space control switching problems in the host controller. The proposed method is to add an approach algorithm to the cruise control at the time of switching from cruise control to space control so that it is switched to space control at the correct switching distance. Through this, the error was improved from 12m error to 4m, and actual vehicle verification was performed.