• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.30 no.3
• /
• pp.46-52
• /
• 2016

When a fault occurs on a transmission line, it is important to identify the fault location as speedily as possible for improvement of the power supply reliability. Generally, distance to fault location is estimated by off line from the recorded data. Conventional fault recorder uses the fault data at one end. This paper deals with the design of an advanced fault recorder for enhancement accuracy of the fault distance estimation and fast detection a fault occurrence position. The major emphasis of the paper will be on the description of the hardware and software of the fault recorder. The fault locator algorithm utilizes a GPS time-synchronized the fault data at both ends. The fault data is transmitted to the other side substation through communication. The advanced fault locator includes a Power module, MPU(Main Processing Unit) module, ADPU(Analog Digital Processing Unit) module, and SIU(Signal Interface Unit) modules. The MMI firmware and software of an advanced fault recording device was implemented.

• International Journal of Contents
• /
• v.15 no.3
• /
• pp.32-38
• /
• 2019

We usually suffer from difficulties in treating or managing Big Data generated from various digital media and/or sensors using traditional mining techniques. Additionally, there are many problems relative to the lack of memory and the burden of the learning curve, etc. in an increasing capacity of large volumes of text when new data are continuously accumulated because we ineffectively analyze total data including data previously analyzed and collected. In this paper, we propose a general-purpose classifier and its structure to solve these problems. We depart from the current feature-reduction methods and introduce a new scheme that only adopts changed elements when new features are partially accumulated in this free-style learning environment. The incremental learning module built from a gradually progressive formation learns only changed parts of data without any re-processing of current accumulations while traditional methods re-learn total data for every adding or changing of data. Additionally, users can freely merge new data with previous data throughout the resource management procedure whenever re-learning is needed. At the end of this paper, we confirm a good performance of this method in data processing based on the Big Data environment throughout an analysis because of its learning efficiency. Also, comparing this algorithm with those of NB and SVM, we can achieve an accuracy of approximately 95% in all three models. We expect that our method will be a viable substitute for high performance and accuracy relative to large computing systems for Big Data analysis using a PC cluster environment.

• Journal of the Korean Wood Science and Technology
• /
• v.49 no.3
• /
• pp.254-266
• /
• 2021

Using forest resources involves solving complex and diverse tasks. At the same time, one of the key goals in the field is improving the quality of forest infrastructure. This direction requires adequate mathematical and economic justification. Moreover, creating an effective infrastructure will not only increase the accessibility and usage volumes of wood and other forest resources, but also contribute to the development of continuous and sustainable forest management. The existing practice of making decisions in terms of the organizational and technological aspects of logging, based on the personal experiences of managers or leading specialists in enterprises, hinders the achievement of constant optimal efficiency. The paper presents results that are a continuation of the research cycle of the authors' team in the fields of optimization and algorithmization of various logging processes. The focus of the study lies in the processing and movement of wood resources, the most valuable products of the investigated groups of enterprises. To this end, the paper presents a developed algorithm for determining an effective technological chain of transportation in logging operations, and for improving loading and unloading processing operations under dynamic natural and production conditions. This algorithm serves as the methodological basis for designing logging infrastructure in a dynamically changing environment.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.4
• /
• pp.693-700
• /
• 2023

Sentiment analysis, a branch of natural language processing that classifies and identifies subjective opinions and emotions in text documents as positive or negative, can be used for various promotions and services through customer preference analysis. To this end, recent research has been conducted utilizing various techniques in machine learning and deep learning. In this study, we propose an optimal language model by comparing the accuracy of sentiment analysis for movie, product, and game reviews using existing RNN-based models and recent Transformer-based language models. In our experiments, LMKorBERT and GPT3 showed relatively good accuracy among the models pre-trained on the Korean corpus.

• Geophysics and Geophysical Exploration
• /
• v.27 no.3
• /
• pp.171-180
• /
• 2024

The processing of seismic data involves analyzing earthquake wave data to understand the internal structure and characteristics of the Earth, which requires high computational power. Recently, machine learning (ML) techniques have been introduced to address these challenges and have been utilized in various tasks such as noise reduction and velocity model construction. However, most studies have focused on specific seismic data processing tasks, limiting the full utilization of similar features and structures inherent in the datasets. In this study, we compared the efficacy of using receiver-wise time-series data ("receiver array") and synchronized receiver signals ("time array") from shotgathers for pretraining a Bidirectional Encoder Representations from Transformers (BERT) model. To this end, shotgather data generated from a synthetic model containing faults was used to perform noise reduction, velocity prediction, and fault detection tasks. In the task of random noise reduction, both the receiver and time arrays showed good performance. However, for tasks requiring the identification of spatial distributions, such as velocity estimation and fault detection, the results from the time array were superior.

• International Journal of Advanced Culture Technology
• /
• v.12 no.3
• /
• pp.67-72
• /
• 2024

The algorithm is applied on the premise that the image generated by AI can be recognized and used smoothly by the user. Other assets are not exposed to the user or discarded because they are unnecessary or unfamiliar. This study aims to expand the scope of the utility of the image generated by AI, which is used as a high-level tool in the design field. To this end, we first examined human information processing and reflection in AI by the cognitive paradigm by examining previous studies and cases, and discussed the value of expansion by focusing on creativity and bottom-up processing of AI's self-organization. Considering the human recogmition process that instinctively grasps an object, the following AI usability was proposed. It is to utilize AI as a high-level tool applied appropriately to human perception, or to utilize the derivative itself by bottom-up self-organization. In addition, it is to set the algorithm to the minimum intervention so that basic elements such as shape, color, size, texture, and movement are composed of figure-ground according to the human perception process that instinctively grasps an object, and to utilize the results. Limiting the use of AI to a tool suitable for human perception and information processing or production by designers or general users is to operate only a part of the convenience and usability of AI. The image creation through AI's self-organization, as seen from the cognitive paradigm, is a step toward opening a new era of design where technical aesthetics meets devices, just as design has been constantly developing in pursuit of novelty and differentiation due to its nature.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.38 no.12
• /
• pp.44-55
• /
• 2001

MAC(Multiply and ACcumulate) is the core operation of multimedia data processing. Because MAC units implemented on traditional DSP units or embedded processors have latency of three cycles and cannot operate on multiple data simultaneously, then, performances are seriously limited. Many high end general purpose microprocessors have SIMD MAC unit as a functional unit. But these high end MAC units must support pipeline structure for various operation modes and high clock frequency, which makes control logic complex and increases chip area. In this paper, a 64bit SIMD MAC unit for embedded processors is designed. It is implemented to have a latency of one clock cycle to remove pipeline control logics and a minimal area overhead for SIMD support is added to existing Booth multipliers.

• Journal of KIISE:Information Networking
• /
• v.36 no.3
• /
• pp.251-255
• /
• 2009

Network coding is a promising technology that increases the system throughput via reducing the number of transmission for a packet delivered from the source node to the destination node. Nevertheless, it suffers from the metrics of end-to-end delay. Network Coding scheme takes more processing delay which occurs as coding node encodes (XOR) a certain number of packets that relayed by the coding node, and more queuing delay which occurs as a packet waits for other packets to be encoded with. Therefore, in this paper, we analyze the dependency of the queuing delay to the arrival rate of each packet. In addition, we analyze and compare the delay in DCF without XOR and DCF with XOR under dynamic traffic.

• KIPS Transactions on Software and Data Engineering
• /
• v.3 no.1
• /
• pp.37-42
• /
• 2014

The abstract should concisely state what was done, how it was done, principal results, and their significance. It should be less than 300 words for all forms of publication. The abstract should be written as one paragraph and should not contain tabular material or numbered references. At the end of abstract, keywords should be given in 3 to 5 words or phrases.

• IE interfaces
• /
• v.20 no.3
• /
• pp.288-297
• /
• 2007

Success of semiconductor/LCD industry depends on its yield and quality of product. For the purpose, FDC (Fault Detection and Classification) system is used to diagnose fault state in main manufacturing processes by monitoring time series data collected by equipment sensors which represent various conditions of the equipment. The data set is segmented at the start and end of each product lot processing by a trigger event module. However, in practice, segmented sensor data usually have the features of data asynchronization such as different start points, end points, and data lengths. Due to the asynchronization problem, false alarm (type I error) and missed alarm (type II error) occur frequently. In this paper, we propose a robust process fault detection system by integrating a process event detection method and a similarity measuring method based on dynamic time warping algorithm. An experiment shows that the proposed system is able to recognize abnormal condition correctly under the asynchronous data situation.