• Journal of Electrical Engineering and Technology
• /
• v.13 no.2
• /
• pp.989-997
• /
• 2018

Massive event stream brings us great challenges in its volume, velocity, variety, value and veracity. Picking up some valuable information from it often faces with long detection time, high memory consumption and low detection efficiency. Aiming to solve the problems above, an efficient complex event detection method based on NFA_HTS (Nondeterministic Finite Automaton_Hash Table Structure) is proposed in this paper. The achievement of this paper lies that we successfully use NFA_HTS to realize the detection of complex event from massive RFID event stream. Specially, in our scheme, after using NFA to capture the related RFID primitive events, we use HTS to store and process the large matched results, as a result, our scheme can effectively solve the problems above existed in current methods by reducing lots of search, storage and computation operations on the basis of taking advantage of the quick classification and storage technologies of hash table structure. The simulation results show that our proposed NFA_HTS scheme in this paper outperforms some general processing methods in reducing detection time, lowering memory consumption and improving event throughput.

• Journal of Korea Multimedia Society
• /
• v.20 no.5
• /
• pp.769-781
• /
• 2017

Deep learning networks like Convolutional Neural Networks (CNNs) show successful performances in many computer vision applications such as image classification, object detection, and so on. For implementation of deep learning networks in embedded system with limited processing power and memory, deep learning network may need to be simplified. However, simplified deep learning network cannot learn every possible scene. One realistic strategy for embedded deep learning network is to construct a simplified deep learning network model optimized for the scene images of the installation place. Then, automatic training will be necessitated for commercialization. In this paper, as an intermediate step toward automatic training under fisheye camera environments, we study more precise human localization in fisheye images, and propose an accurate human localization method, Automatic Ground-Truth Labelling Method (AGTLM). AGTLM first localizes candidate human object bounding boxes by utilizing GoogLeNet-LSTM approach, and after reassurance process by GoogLeNet-based CNN network, finally refines them more correctly and precisely(tightly) by applying saliency object detection technique. The performance improvement of the proposed human localization method, AGTLM with respect to accuracy and tightness is shown through several experiments.

• Smart Structures and Systems
• /
• v.24 no.1
• /
• pp.111-125
• /
• 2019

Metallic shape memory alloys present fascinating physical properties such as their super-elastic behavior in austenite phase, which can be exploited for providing a structure with both a self-centering capability and an increased ductility. More or less accurate numerical models have been introduced to model their behavior along the last 25 years. This is the reason for which the literature is rich of suggestions/proposals on how to implement this material in devices for passive and semi-active control. Nevertheless, the thermo-mechanical coupling characterizing the first-order martensite phase transformation process results in several macroscopic features affecting the alloy performance. In particular, the effects of day-night and winter-summer temperature excursions require special attention. This aspect might imply that the deployment of some devices should be restricted to indoor solutions. A further aspect is the dependence of the behavior from the geometry one adopts. Two fundamental lacks of symmetry should also be carefully considered when implementing a SMA-based application: the behavior in tension is different from that in compression, and the heating is easy and fast whereas the cooling is not. This manuscript focuses on the passive devices recently proposed in the literature for civil engineering applications. Based on the challenges above identified, their actual feasibility is investigated in detail and their long term performance is discussed with reference to their fatigue life. A few available semi-active solutions are also considered.

• The Korean Journal of Financial Management
• /
• v.21 no.1
• /
• pp.149-181
• /
• 2004

In this study we perform empirical tests using KOSPI return to investigate the existence of nonlinear characteristics in the generating process of stock returns. There are three categories in empirical tests; the test of nonlinear dependence, nonlinear stochastic process and nonlinear deterministic chaos. According to the analysis of nonlinearity, stock returns are not normally distributed but leptokurtic, and appear to have nonlinear dependence. And it's decided that the nonlinear structure of stock returns can not be completely explained using nonlinear stochastic models of ARCH-type. Nonlinear deterministic chaos system is the feedback system, which the past incidents influence the present, and it is the fractal structure with self-similarity and has the sensitive dependence on initial conditions. To summarize the results of chaos analysis for KOSPI return, it is the persistent time series, which is not IID and has long memory, takes biased random walk, and is estimated to be fractal distribution. Also correlation dimension, as the approximation of fractal dimension, converged stably within 3 and 4, and maximum Lyapunov exponent has positive value. This suggests that chaotic attractor and the sensitive dependence on initial conditions exist in stock returns. These results fit into the characteristics of chaos system. Therefore it's decided that the generating process of stock returns has nonlinear deterministic structure and follow chaotic process.

• Korean Journal of Remote Sensing
• /
• v.35 no.5_1
• /
• pp.681-692
• /
• 2019

In this paper, a hybrid deep learning model, called 2D convolution with bidirectional long short-term memory (2DCBLSTM), is presented that can effectively combine both spatial and temporal features for crop classification. In the proposed model, 2D convolution operators are first applied to extract spatial features of crops and the extracted spatial features are then used as inputs for a bidirectional LSTM model that can effectively process temporal features. To evaluate the classification performance of the proposed model, a case study of crop classification was carried out using multi-temporal unmanned aerial vehicle images acquired in Anbandegi, Korea. For comparison purposes, we applied conventional deep learning models including two-dimensional convolutional neural network (CNN) using spatial features, LSTM using temporal features, and three-dimensional CNN using spatio-temporal features. Through the impact analysis of hyper-parameters on the classification performance, the use of both spatial and temporal features greatly reduced misclassification patterns of crops and the proposed hybrid model showed the best classification accuracy, compared to the conventional deep learning models that considered either spatial features or temporal features. Therefore, it is expected that the proposed model can be effectively applied to crop classification owing to its ability to consider spatio-temporal features of crops.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.10a
• /
• pp.501-503
• /
• 2017

In the water sector, various data such as flow rate, pressure, water quality and water level are collected during the whole process of water purification plant and piping system. The collected data is stored in each water treatment plant's DB, and the collected data are combined in the regional DB and finally stored in the database server of the head office of the Korea Water Resources Corporation. Various abnormal data can be generated when a measuring instrument measures data or data is communicated over various processes, and it can be classified into missing data and wrong data. The cause of each abnormal data is different. Therefore, there is a difference in the method of detecting the wrong side and the missing side data, but the method of cleansing the data is the same. In this study, a program that can automatically refine missing or wrong data by applying deep learning LSTM (Long Short Term Memory) algorithm will be studied.

• Smart Media Journal
• /
• v.12 no.11
• /
• pp.103-112
• /
• 2023

Recently, the need to prevent battery fires and accidents has emerged, as the use of lithium-ion batteries has increased. In order to prevent accidents, it is necessary to predict the state of health (SOH) and check the replacement timing of the battery with a lot of degradation. This paper proposes a model for predicting the degradation state of a battery by using four battery degradation indicators: maximum voltage arrival time, current change time, maximum temperature arrival time, and incremental capacity (IC) that can be obtained in the battery charging process, and LSTM using an attention mechanism. The performance of the proposed model was measured using the NASA battery data set, and the predictive performance was improved compared to that of the general LSTM model, especially in the SOH 90-70% section, which is close to the battery replacement cycle.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.47 no.2
• /
• pp.155-167
• /
• 2024

Distribution and logistics industries contribute some of the biggest GDP(gross domestic product) in South Korea and the number of related companies are quarter of the total number of industries in the country. The number of retail tech companies are quickly increased due to the acceleration of the online and untact shopping trend. Furthermore, major distribution and logistics companies try to achieve integrated data management with the fulfillment process. In contrast, small and medium distribution companies still lack of the capacity and ability to develop digital innovation and smartization. Therefore, in this paper, a deep learning-based demand forecasting & recommendation model is proposed to improve business competitiveness. The proposed model is developed based on real sales transaction data to predict future demand for each product. The proposed model consists of six deep learning models, which are MLP(multi-layers perception), CNN(convolution neural network), RNN(recurrent neural network), LSTM(long short term memory), Conv1D-BiLSTM(convolution-long short term memory) for demand forecasting and collaborative filtering for the recommendation. Each model provides the best prediction result for each product and recommendation model can recommend best sales product among companies own sales list as well as competitor's item list. The proposed demand forecasting model is expected to improve the competitiveness of the small and medium-sized distribution and logistics industry.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2000.10a
• /
• pp.124-127
• /
• 2000

The finite element analyses of the wrinkling initiation and growth in the sheet metal forming process provide the detailed information about the wrinkling behavior of sheet metal. The direct analyses of the wrinkling initiation and growth, however, bring about a little difficulty in complex industrial problems because it needs large memory size and long computation time. For the description of wrinkling growth, the mesh elements should be sufficiently small and the size of finite element matrix becomes large. In the static implicit finite element method therefore, the direct analysis of wrinkling growth in a complex sheet metal forming process is rather difficult. From the industrial viewpoint of tooling design, the readily available information of possibility and location of wrinkling is sometimes more preferable to the detailed time-consuming information. In the present study, therefore, the wrinkling factor that shows locations and relative possibility of wrinkling initiation is proposed as a convenient tool of relative wrinkling estimation based on the energy criterion. The location and relative possibility of wrinkling initiation are predicted by calculating the wrinkling factor in various sheet metal forming processes such as cylindrical cup deep drawing, spherical cup deep drawing, and elliptical cup deep drawing. The wrinkling factor is also implemented in the analysis of the door inner stamping process to predict wrinkling.

• Transactions of Materials Processing
• /
• v.9 no.6
• /
• pp.653-662
• /
• 2000

Wrinkling is one of the major defects in sheet metal products together with tearing, springback and other geometric and surface defects. The initiation and growth of wrinkles are influenced by many factors such as stress ratios, mechanical properties of the sheet material, geometry of the workpiece, contact condition, etc. It is difficult to analyze the wrinkling initiation and growth considering all the factors because the effects of the factors are very complex and the wrinkling behavior may show a wide scatter of data even for small deviations of factors. The finite element analyses of the wrinkling initiation and growth in the sheet metal forming process provide the detailed information about the wrinkling behavior of sheet metal. The direct analyses of the wrinkling initiation and growth, however, bring about a little difficulty in complex industrial problems because it needs large memory size and long computation time. In the present study, therefore, a global-local analysis technique is introduced for the computational efficiency. Through the analysis of wrinkling in the door inner stamping process, the efficiency of the global-local analysis technique is investigated.