• Korean Journal of Applied Biomechanics
• /
• v.32 no.1
• /
• pp.1-8
• /
• 2022

Objective: This study proposes a methodology to analyze important variables that have a significant impact on the putting direction prediction using a machine learning-based putting direction prediction model trained with IMU sensor data. Method: Putting data were collected using an IMU sensor measuring 12 variables from 6 adult males in their 20s at K University who had no golf experience. The data was preprocessed so that it could be applied to machine learning, and a model was built using five machine learning algorithms. Finally, by comparing the performance of the built models, the model with the highest performance was selected as the proposed model, and then 12 variables of the IMU sensor were applied one by one to analyze important variables affecting the learning performance. Results: As a result of comparing the performance of five machine learning algorithms (K-NN, Naive Bayes, Decision Tree, Random Forest, and Light GBM), the prediction accuracy of the Light GBM-based prediction model was higher than that of other algorithms. Using the Light GBM algorithm, which had excellent performance, an experiment was performed to rank the importance of variables that affect the direction prediction of the model. Conclusion: Among the five machine learning algorithms, the algorithm that best predicts the putting direction was the Light GBM algorithm. When the model predicted the putting direction, the variable that had the greatest influence was the left-right inclination (Roll).

• The Journal of Korea Robotics Society
• /
• v.7 no.3
• /
• pp.205-215
• /
• 2012

A GPS sensor is widely used in many areas such as navigation, or air traffic control. Particularly, the car navigation system is equipped with GPS sensor for locational information. However, when a car goes through a tunnel, forest, or built-up area, GPS receiver cannot get the enough number of satellite signals. In these situations, a GPS receiver does not reliably work. A GPS error can be formulated by sum of bias error and sensor noise. The bias error is generated by the geometric arrangement of satellites and sensor noise error is generated by the corrupted signal noise of receiver. To enhance GPS sensor accuracy, these two kinds of errors have to be removed. In this research, we make the road database which includes Road Database File (RDF). RDF includes road information such as road connection, road condition, coordinates of roads, lanes, and stop lines. Among the information, we use the stop line coordinates as a feature point to correct the GPS bias error. If the relative distance and angle of a stop line from a car are detected and the detected stop line can be associated with one of the stop lines in the database, we can measure the bias error and correct the car's location. To remove the other GPS error, sensor noise, the Kalman filter algorithm is used. Additionally, using the RDF, we can get the information of the road where the car belongs. It can be used to help the GPS correction algorithm or to give useful information to users.

• Smart Structures and Systems
• /
• v.20 no.2
• /
• pp.163-173
• /
• 2017

Structural health monitoring (SHM) of civil infrastructure using fiber Bragg grating sensor networks (FBGSNs) has received significant public attention in recent years. However, there is currently little research on the health-monitoring technology of high-piled wharfs in coastal ports using the fiber Bragg grating (FBG) sensor technique. The benefits of FBG sensors are their small size, light weight, lack of conductivity, resistance corrosion, multiplexing ability and immunity to electromagnetic interference. Based on the properties of high-piled wharfs in coastal ports and servicing seawater environment and the benefits of FBG sensors, the SHM system for a high-piled wharf in the Tianjin Port of China is devised and deployed partly using the FBG sensor technique. In addition, the health-monitoring parameters are proposed. The system can monitor the structural mechanical properties and durability, which provides a state-of-the-art mean to monitor the health conditions of the wharf and display the monitored data with the BIM technique. In total, 289 FBG stain sensors, 87 FBG temperature sensors, 20 FBG obliquity sensors, 16 FBG pressure sensors, 8 FBG acceleration sensors and 4 anode ladders are installed in the components of the back platform and front platform. After the installation of some components in the wharf construction site, the good signal that each sensor measures demonstrates the suitability of the sensor setup methods, and it is proper for the full-scale, continuous, autonomous SHM deployment for the high-piled wharf in the costal port. The South 27# Wharf SHM system constitutes the largest deployment of FBG sensors for wharf structures in costal ports to date. This deployment demonstrates the strong potential of FBGSNs to monitor the health of large-scale coastal wharf structures. This study can provide a reference to the long-term health-monitoring system deployment for high-piled wharf structures in coastal ports.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.12 no.5
• /
• pp.161-167
• /
• 2012

Portable devices such as smartphones with built-in wireless LAN to the prevalence of anyone using. But the wireless Internet connection and positioning services are limited to high-quality wireless service, they may not be available. Thus, wireless LAN infrared sensor in the same way as with angry alternative way wireless capabilities of the application automatically identify the location of the Sensor application as an alternative method is proposed. Thus, wireless LAN, such as infrared sensors and other alternzative methods of wireless features in a way where the application can recognize and automatically recognize the sensor application as an alternative method is proposed. Sensor is signals between wireless LAN and access points using the sensitivity, WLAN antenna with omni-directional signal output operation of the sensor is assumed to be recognize this by putting a direction to obtain through the proposed algorithm, Sensors such as photo-coupler without direct recognition sensor, wireless LAN and access points, the same function as the connection between the sensitivity to perform its function was to utilizing.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.12
• /
• pp.2885-2890
• /
• 2014

Localization of a sensor network node using machine learning has been recently studied. It is easy for Support vector machines algorithm to implement in high level language enabling parallelism. Raspberrypi is a linux system which can be used as a sensor node. Pi can be used to construct IP based Hadoop clusters. In this paper, we realized Support vector machine using python language and built a sensor network cluster with 5 Pi's. We also established a Hadoop software framework to employ MapReduce mechanism. In our experiment, we implemented the test sensor network with a variety of parameters and examined based on proficiency, resource evaluation, and processing time. The experimentation showed that with more execution power and memory volume, Pi could be appropriate for a member node of the cluster, accomplishing precise classification for sensor localization using machine learning.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2003.06a
• /
• pp.1332-1335
• /
• 2003

This paper presents the enhancement of oxygen transfer efficiency using the vibrating intravascular lung assist device (VIVLAD) in in-vitro experiments for patients having chronic respiratory problems. The test section was a cylinder duct with the inner diameter of 30 mm. The flow rate was controlled by the pump and monitored by a built-in flow meter. The vibration apparatus was composed of a piezo-vibrator, a function generator. and a power amplifier. The direction of vibration was radial to the fluid flow. Gas flow rates of up to 6 l/min through the 120-cm-Jong hollow fibers have been achieved by exciting a piezo-vibrator. The output of PVDF sensor were investigated by various frequencies in VIVLAD. The experimental results showed that VIVLAD would be enhance oxygen transfer efficiency.

• Proceedings of the KSR Conference
• /
• 2007.11a
• /
• pp.989-993
• /
• 2007

It was intended in this study to seek for the measures to utilize the USN technique, which has high usability due to low price and low power consumption, in air quality monitoring. As a method, the sensors of temperature, humidity, particulate matters (PM10), and carbon dioxide ($CO_2$) were installed in the self-manufactured sensor nodes; the nodes were installed in the waiting rooms and platforms of a subway station and the measurements were collected at real time with use of a computer which micro gateway was built in. Collected data was to be processed by the statistics program installed in the computer; the collected data is to be used in managing the air quality of stations after transmission to the ventilation system of ventilation chambers.

• Journal of Magnetics
• /
• v.3 no.3
• /
• pp.82-85
• /
• 1998

A portable HTS RF SQUID-based system, weighing less than 20 kg has been built for susceptometry applications in weak magnetic fields, It includes a YBCO sensor for measuring the axial magnetic field component with a resolution of about $7{\times}10^{-13} T/Hz^{1/2}.$ This is determined by the intrinsic magnetic noise in the quasi-white noise region. There is a relaxation for a sudden increase in field due to magnetic flux creep in HTS. In this instance the time did not exceed 3~5 minutes.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.3
• /
• pp.39-47
• /
• 2015

In this paper, we propose a new time synchronization algorithm using CoAP(constrained-application protocol) in sensor network environment, which handles a technique that synchronizes an explicit timestamp between sensor nodes not including an additional module for time-setting and sensor node gateway linked to internet time server. CoAP is a standard protocol for sensor data communication among sensor nodes and sensor node gateway to be built much less memory and power supply in constrained network surroundings including serious network jitter, packet losses, etc. We have supplied an exact time synchronization implementation among small and cheap IP-based sensor nodes or non-IP based sensor nodes and sensor node gateway in sensor network using CoAP message header's option extension. On behalf of conventional network time synchronization method, as our approach uses an exclusive protocol 'CoAP' in sensor network, it is not to become an additional burden for synchronization service to sensor nodes or sensor node gateway. This method has an average error about 2ms comparing to NTP service and offers a low-cost and robust network time synchronization algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.2
• /
• pp.129-138
• /
• 2009

While the issue of caring for the elderly that faces the modern society has reached a serious level, it is expected that it will be particularly true in the Republic of Korea where an aged, not an aging, society is impending. In this paper I did research on establishing behavioral model of residents who dwell in home or welfare facilities. I suggested a behavioral model in daily living, $W_{ip}(n)$, based on event triggering. A multi-hop routing-based wireless luminance/temperature sensor network was built based on the proposed resident's behavioral model. 1 did experiments on behavioral activities of residents on the wireless sensor network system. According to experimental results. I could classify whether the daily activity of a resident someday is regular or not. These experimental results show that the proposed behavioral model is highly applicable in caring for residents in home or welfare facilities effectively in the future.