• Journal of the Korea Society of Computer and Information
• /
• v.27 no.2
• /
• pp.25-32
• /
• 2022

In this paper, we propose a pet monitoring system based on deep learning using an activity recognition device. The system consists of a pet's activity recognition device, a pet owner's smart device, and a server. Accelerometer and gyroscope data were collected from an Arduino-based activity recognition device, and the number of steps was calculated. The collected data is pre-processed and the amount of activity is measured by recognizing the activity in five types (sitting, standing, lying, walking, running) through a deep learning model that hybridizes CNN and LSTM. Finally, monitoring of changes in the activity, such as daily and weekly briefing charts, is provided on the pet owner's smart device. As a result of the performance evaluation, it was confirmed that specific activity recognition and activity measurement of pets were possible. Abnormal behavior detection of pets and expansion of health care services can be expected through data accumulation in the future.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.4
• /
• pp.1224-1248
• /
• 2022

In various sensor network applications, such as climate observation organizations, sensor nodes need to collect information from time to time and pass it on to the recipient of information through multiple bounces. According to field tests, this information corresponds to most of the energy use of the sensor hub. Decreasing the measurement of information transmission in sensor networks becomes an important issue.Compression sensing (CS) can reduce the amount of information delivered to the network and reduce traffic load. However, the total number of classification of information delivered using pure CS is still enormous. The hybrid technique for utilizing CS was proposed to diminish the quantity of transmissions in sensor networks.Further the energy productivity is a test task for the sensor nodes. However, in previous studies, a clustering approach using hybrid CS for a sensor network and an explanatory model was used to investigate the relationship between beam size and number of transmissions of hybrid CS technology. It uses efficient data integration techniques for large networks, but leads to clone attacks or attacks. Here, a new algorithm called SBEA (Snowball Endurance Algorithm) was proposed and tested with a bow. Thus, you can extend the battery life of your WSN by running effective copy detection. Often, multiple nodes, called observers, are selected to verify the reliability of the nodes within the network. Personal data from the source centre (e.g. personality and geographical data) is provided to the observer at the optional witness stage. The trust and reputation system is used to find the reliability of data aggregation across the cluster head and cluster nodes. It is also possible to obtain a mechanism to perform sleep and standby procedures to improve the life of the sensor node. The sniffers have been implemented to monitor the energy of the sensor nodes periodically in the sink. The proposed algorithm SBEA (Snowball Endurance Algorithm) is a combination of ERCD protocol and a combined mobility and routing algorithm that can identify the cluster head and adjacent cluster head nodes.This algorithm is used to yield the network life time and the performance of the sensor nodes can be increased.

• Steel and Composite Structures
• /
• v.44 no.2
• /
• pp.199-210
• /
• 2022

This study build finite element analysis (FEA) models describing the bending events of coiled tubing (CT) at the wellhead and trips into the hole, accurately provide the state of stress and strain while the CT is in service. The bending moment and axial force history curves are used as loads and boundary conditions in the diametrical growth models to ensure consistency with the actual working conditions in field operations. The simulation diametrical growth results in this study are more accurate and reasonable. Analysis the factors influencing fatigue and diametrical growth shows that the internal pressure has a first-order influence on fatigue, followed by the radius of the guide arch, reel and the CT diameter. As the number of trip cycles increase, fatigue damage, residual stress and strain cumulatively increase, until CT failure occurs. Significant residual stresses remain in the CT cross-section, and the CT exhibits a residual curvature, the initial residual bending configuration of CT under wellbore constraints, after running into the hole, is sinusoidal. The residual stresses and residual bending configuration significantly decrease the buckling load, making the buckling and buckling release of CT in the downhole an elastic-plastic process, exacerbating the helical lockup. The conclusions drawn in this study will improve CT models and contribute to the operational and economic success of CT services.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.2
• /
• pp.61-66
• /
• 2023

As the use of smartphones expands to various areas and the level of multitasking increases, the support of swap is becoming increasingly important. However, swap support in smartphones is known to cause excessive storage traffic, resulting in memory thrashing. In this paper, we analyze how the thrashing of swaps that occurred in early smartphones has changed with the advancement of smartphone hardware. As a result of this analysis, we show that the swap thrashing problem can be resolved to some extent when the memory size increases. However, we also show that thrashing still occurs when the number of running apps continues to increase. Based on further analysis, we observe that this thrashing is caused by some hot data and suggest a way to solve this through an NVM-based architecture. Specifically, we show that a small size NVM with judicious management can resolve the performance degradation caused by smartphone swap.

• Journal of the Society of Naval Architects of Korea
• /
• v.60 no.2
• /
• pp.95-109
• /
• 2023

In the 4th industrial revolution, changes in the technological paradigm have had a direct impact on the maintenance system of ships. The 2-stroke low speed engine system integrates with the core equipment required for propulsive power. The Condition Based Management (CBM) is defined as a technology that predictive maintenance methods in existing calender-based or running time based maintenance systems by monitoring the condition of machinery and diagnosis/prognosis failures. In this study, we have established a framework for CBM technology development on our own, and are engaged in engineering-based failure analysis, data development and management, data feature analysis and pre-processing, and verified the reliability of failure mode DB using LSTM algorithms. We developed various simulated failure mode scenarios for 2-stroke low speed engine and researched to produce data on onshore basis test_beds. The analysis and pre-processing of normal and abnormal status data acquired through failure mode simulation experiment used various Exploratory Data Analysis (EDA) techniques to feature extract not only data on the performance and efficiency of 2-stroke low speed engine but also key feature data using multivariate statistical analysis. In addition, by developing an LSTM classification algorithm, we tried to verify the reliability of various failure mode data with time-series characteristics.

• International Journal of Computer Science & Network Security
• /
• v.23 no.1
• /
• pp.46-52
• /
• 2023

With billions of IoT (Internet of Things) devices populating various emerging applications across the world, detecting anomalies on these devices has become incredibly important. Advanced Intrusion Detection Systems (IDS) are trained to detect abnormal network traffic, and Machine Learning (ML) algorithms are used to create detection models. In this paper, the NSL-KDD dataset was adopted to comparatively study the performance and efficiency of IoT anomaly detection models. The dataset was developed for various research purposes and is especially useful for anomaly detection. This data was used with typical machine learning algorithms including eXtreme Gradient Boosting (XGBoost), Support Vector Machines (SVM), and Deep Convolutional Neural Networks (DCNN) to identify and classify any anomalies present within the IoT applications. Our research results show that the XGBoost algorithm outperformed both the SVM and DCNN algorithms achieving the highest accuracy. In our research, each algorithm was assessed based on accuracy, precision, recall, and F1 score. Furthermore, we obtained interesting results on the execution time taken for each algorithm when running the anomaly detection. Precisely, the XGBoost algorithm was 425.53% faster when compared to the SVM algorithm and 2,075.49% faster than the DCNN algorithm. According to our experimental testing, XGBoost is the most accurate and efficient method.

• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.452-459
• /
• 2023

The core power control is an important issue for the study of dynamic characteristics in China initiative accelerator driven subcritical system (CiADS), which has direct impact on the control strategy and safety analysis process. The CiADS is an experimental facility that is only controlled by the proton beam intensity without considering the control rods in the current engineering design stage. In order to get the optimized operation scheme with the stable and reliable features, the variation of beam intensity using the continuous and periodic control approaches has been adopted, and the change of collimator and the adjusting of duty ratio have been proposed in the power control process. Considering the neutronics and the thermal-hydraulics characteristics in CiADS, the physical model for the core power control has been established by means of the point reactor kinetics method and the lumped parameter method. Moreover, the multi-inputs single-output (MISO) logical structure for the power control process has been constructed using proportional integral derivative (PID) controller, and the meta-heuristic algorithm has been employed to obtain the global optimized parameters for the stable running mode without producing large perturbations. Finally, the verification and validation of the control method have been tested based on the reference scenarios in considering the disturbances of spallation neutron source and inlet temperature respectively, where all the numerical results reveal that the optimization method has satisfactory performance in the CiADS core power control scenarios.

• Journal of Korean Society for Quality Management
• /
• v.50 no.1
• /
• pp.153-167
• /
• 2022

Purpose: The purpose of this study is to guide the operation managers who plan daily production of large mass-processing facility that services multi-customers with multi-product, small-batch item characteristics by providing the practical best production quantity and the inventory allowed to build. Methods: Close observation of a subcontract paint-shop operator captured the daily decision process which was reflected in the subcontractor-unique mathematical model and the system dynamics simulation model. Multiple simulations were run to find the practical best production quantity and the maximum allowable stock level of inventory that did not undermine the profit from practical best daily production. Actual data and a few constant values were obtained from the firm under study. Results: While the inventory holding cost for the customer-owned material harms the total profit of the subcontractor, the running cost of the processing facility hinders production in small batches. This balances the maximum possible productions and results in practical best daily production which can be found through simulation runs with actual data. The maximum level of stocked inventory is deduced from the practical best daily production. Conclusion: To build a large volume that enables economy-of-scale production, operators should deal with multi-product small-batch items from multiple customers. When the planned schedule of the time and amount of material in-flow tend not to be reliable, operators can find it practical to execute level production across the planning horizon instead of adjusting to day-to-day in-flow fluctuations.

• Journal of Auto-vehicle Safety Association
• /
• v.15 no.2
• /
• pp.35-41
• /
• 2023

On December 19, 2015, as Article 29-3 (Installation of Accident Recording Devices and Provision of Information) of Motor Vehicle Management Act came into force, In Korea, the EDR (Event Data Recorder) reports are often used for the analysis of various traffic accident cases such as multiple collisions, traffic insurance crimes, and sudden unintended acceleration (SUA), and the others. So many investigators have analyzed the driver's behavior and vehicle situation by comparing the time zero in the EDR report to the actual crash time in dash-cam (or CCTV). Time zero (T0) is defined as the reference time for the record interval or time interval when recording an accident in Article 56-2, Enforcement rule of Performance and Standard for Automobile and Automotive parts. Also in the EDR report, time zero (T0) is defined as whichever of the following occurs first; 1. "wake-up" by an air-bag control system, 2. Continuously running algorithms (by monitoring of longitudinal or lateral delta-V), 3. Deployment of a non-reversible deployment restraint. We have already proposed the "Flowchart & Checklist" to adopt the EDR report for traffic accident investigation and the necessity of specialized institutions or courses to systematically educate or analyze the EDR data. Therefore, in this paper, we report to traffic accident investigators notable points and analysis methods based on some real-world traffic accidents that can be misjudged in specifying time zero (T0).

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.6
• /
• pp.519-524
• /
• 2023

The study summarizes the experience of operation in the application of flipped learning to various IT-related liberal arts subjects, and proposes a specific application method. So far, most of the studies have analyzed various strategies and learner responses to flipped learning. Currently, it is the time when teachers, who are the main operators of the flipped learning class, need to study how to provide immediate feedback and application while running the relevant courses. Studies related to this are gradually coming out. In general, most of the studies on sharing reference materials through the results after applying various strategies such as developing the structure of class operation by instructors themselves, combining them with discussion classes, or developing various contents. This study proposes a method to analyze how various strategies can be applied in the subject and obtain results simultaneously with class operation by analyzing unstructured data, which is a survey that can receive immediate feedback.