• Environmental Engineering Research
• /
• v.15 no.2
• /
• pp.123-126
• /
• 2010

Modeling non-point pollution across multiple scales has become an important environmental issue. As a more representative and practical approach in quantifying and qualifying surface water, a modular neural network (MNN) was implemented in this study. Two different site-scales ($1.5\;{\times}\;10^5$ and $1.62\;{\times}\;10^6\;m^2$) with the same plants, soils, and paddy field management practices, were selected. Hydrologic data (rainfall, irrigation and surface discharge) and water quality data (time-series nutrient loadings) were continuously monitored and then used for the verification of MNN performance. Correlation coefficients (R) for the results predicted from the networks versus measured values were within the range of 0.41 to 0.95. The small block could be extrapolated to the large field for the rainfall-surface drainage process. Nutrient prediction produced less favorable results due to the complex phenomena of nutrients in the drainage water. However, the feasibility of using MNN to generate improved prediction accuracy was demonstrated if more hydrologic and environmental data are provided. The study findings confirmed the estimation accuracy of the upscaling from a small-segment block to large-scale paddy field, thereby contributing to the establishment of water quality management for sustainable agriculture.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.11
• /
• pp.820-829
• /
• 2006

Wireless sensor networks consisting of hundreds to thousands of nodes are expected to be increasingly deployed in coming years, as they enable reliable monitoring and analysis of physical worlds. These networks have unique features that are very different from traditional networks, e.g., the numerous numbers of nodes, limitation in power, processing, and memory. Due to these unique features of wireless sensor networks, sensor data management including querying becomes a challenging problem. Furthermore, due to wide popularization of the Internet and its facility in use, it is generally accepted that an unattended network can be efficiently managed and monitored over the Internet. In particular, in order to more efficiently query and manage data in a sensor network. in this paper, the architecture of a sensor gateway including web-based query server is presented and its implementation detail is illustrated. The presented web-based gateway is largely divided into two important parts: Internet part and sensor network part. The sensor network part plays an important role of handling a variety of sensor networks, including flat or hierarchical network architecture, by using internally layered architecture for efficiently querying and managing data in a sensor network. In addition, the Internet part provides a modular gateway function for favorable exchange between the sensor network and Internet.

• Journal of KIISE:Software and Applications
• /
• v.41 no.7
• /
• pp.481-491
• /
• 2014

The power production using renewable energy is more important because of a limited amount of fossil fuel and the problem of global warming. A concentrative photovoltaic system comes into the spotlight with high energy production, since the rate of power production using solar energy is proliferated. These systems, however, need to sophisticated tracking methods to give the high power production. In this paper, we propose a hierarchical tracking system using modular Bayesian networks and a naive Bayes classifier. The Bayesian networks can respond flexibly in uncertain situations and can be designed by domain knowledge even when the data are not enough. Bayesian network modules infer the weather states which are classified into nine classes. Then, naive Bayes classifier selects the most effective method considering inferred weather states and the system makes a decision using the rules. We collected real weather data for the experiments and the average accuracy of the proposed method is 93.9%. In addition, comparing the photovoltaic efficiency with the pinhole camera system results in improved performance of about 16.58%.

• Proceedings of the IEEK Conference
• /
• 2005.11a
• /
• pp.1023-1026
• /
• 2005

In the ubiquitous computing environment, an intelligent vehicle is defined as a sensor node with a capability of intelligence and communication in a wire and wireless network space. To make it real, a lot of problems should be addressed in the aspect of vehicle mobility, in-vehicle communication, common service platform and the connection of heterogeneous networks to provide a driver with several intelligent information services beyond the time and space. In this paper, we present an intelligent information system for managing in-vehicle sensor network and a vehicle gateway for connecting the external networks. The in-vehicle sensor network connected with several sensor nodes is used to collect sensor data and control the vehicle based on CAN protocol. Each sensor node is equipped with a reusable modular node architecture, which contains a common CAN stack, a message manager and an event handler. The vehicle gateway makes vehicle control and diagnosis from a remote host possible by connecting the in-vehicle sensor network with an external network. Specifically, it gives an access to the external mobile communication network such as CDMA. Some experiments was made to find out how long it takes to communicate between a vehicle's intelligent information system and an external server in the various environment. The results show that the average response time amounts to 776ms at fixed place, 707ms at rural area and 910ms at urban area.

• Journal of KIISE
• /
• v.42 no.3
• /
• pp.320-327
• /
• 2015

Although service robots in various fields are being commercialized, most of them have problems that depend on explicit commands by users and have difficulty to generate robust reactions of the robot in the unstable condition using insufficient sensor data. To solve these problems, we modeled mirror neuron and theory of mind systems, and applied them to a robot agent to show the usefulness. In order to implement quick and intuitive response of the mirror neuron, the proposed intention-response model utilized behavior selection networks considering external stimuli and a goal, and in order to perform reactions based on the long-term action plan of theory of mind system, we planned behaviors of the sub-goal unit using a hierarchical task network planning, and controled behavior selection network modules. Experiments with various scenarios revealed that appropriate reactions were generated according to external stimuli.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2000.11a
• /
• pp.56-58
• /
• 2000

Post-genomics may be defined in different ways depending on how one views the challenges after the genome. A popular view is to follow the concept of the central dogma in molecular biology, namely from genome to transcriptome to proteome. Projects are going on to analyze gene expression profiles both at the mRNA and protein levels and to catalog protein 3D structure families, which will no doubt help the understanding of information in the genome. However complete, such catalogs of genes, RNAs, and proteins only tell us about the building blocks of life. They do not tell us much about the wiring (interaction) of building blocks, which is essential for uncovering systemic functional behaviors of the cell or the organism. Thus, an alternative view of post-genomics is to go up from the molecular level to the cellular level, and to understand, what I call, the "interactome"or a complete picture of molecular interactions in the cell. KEGG (http://www.genome.ad.jp/kegg/) is our attempt to computerize current knowledge on various cellular processes as a collection of "generalized"protein-protein interaction networks, to develop new graph-based algorithms for predicting such networks from the genome information, and to actually reconstruct the interactomes for all the completely sequenced genomes and some partial genomes. During the reconstruction process, it becomes readily apparent that certain pathways and molecular complexes are present or absent in each organism, indicating modular structures of the interactome. In addition, the reconstruction uncovers missing components in an otherwise complete pathway or complex, which may result from misannotation of the genome or misrepresentation of the KEGG pathway. When combined with additional experimental data on protein-protein interactions, such as by yeast two-hybrid systems, the reconstruction possibly uncovers unknown partners for a particular pathway or complex. Thus, the reconstruction is tightly coupled with the annotation of individual genes, which is maintained in the GENES database in KEGG. We are also trying to expand our literature surrey to include in the GENES database most up-to-date information about gene functions.

• Genomics & Informatics
• /
• v.2 no.4
• /
• pp.191-194
• /
• 2004

Exponentially increasing biopathway data in recent years provide us with means to elucidate the large-scale modular organization of the cell. Given the existing information on metabolic and regulatory networks, inferring biopathway information through scientific reasoning or data mining of large scale array data or proteomics data get great attention. Naturally, there is a need for a user-friendly system allowing the user to combine large and diverse pathway data sets from different resources. We built a data warehouse - BIOWAY - for analyzing and visualizing biological pathways, by integrating and customizing resources. We have collected many different types of data in regards to pathway information, including metabolic pathway data from KEGG/LIGAND, signaling pathway data from BIND, and protein information data from SWISS-PROT. In addition to providing general data retrieval mechanism, a successful user interface should provide convenient visualization mechanism since biological pathway data is difficult to conceptualize without graphical representations. Still, the visual interface in the previous systems, at best, uses static images only for the specific categorized pathways. Thus, it is difficult to cope with more complex pathways. In the BIOWAY system, all the pathway data can be displayed in computer generated graphical networks, rather than manually drawn image data. Furthermore, it is designed in such a way that all the pathway maps can be expanded or shrinked, by introducing the concept of super node. A subtle graphic layout algorithm has been applied to best display the pathway data.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.05a
• /
• pp.256-259
• /
• 2002

Recently, most vehicles has the Automatic transmission system as their transmission system. The automatic transmission system operates with fixed shift patterns. In the opposite of manual operation, it is easy and convenient for driving. Though these merit, the system can not evaluate the driver's intension because of usage of firmed shift pattern. Especially, when the load has declination the AT system must operate for engine break effect. Namely, if the vehicle drives on the load of decrease, the acceleration of the vehicle goes to high then. At that time, the shift goes to down position the vehicle has some negative acceleration with the resistance of engine. To consider driver's intension in this case, we must consider both the driving intensity of driver and the status of load. In this paper, we developed flexible automatic transmission system by using the proposed moduled neural networks which can learn the status of the load and driver's intensity As a result, we compare the transmission system using firmed shift pattern and the proposed transmission system and show the good performance in the change of shift position.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.10a
• /
• pp.469-474
• /
• 1997

To provide tile information ell severe accident progression is very important for advanced or new type of nuclear power plant (NPP) design. A parametric study, therefore was performed to investigate the effect of thermal hydraulic design parameters ell severe accident progression of pressurized water reactors (PWRs), Nine parameters, which are considered important in NPP design or severe accident progression, were selected among the various thermal hydraulic design parameters. The backpropagation neural network (BPN) was used to determine parameters, which might more strongly affect the severe accident progression, among mile parameters. For training. different input patterns were generated by the latin hypercube sampling (LHS) technique and then different target patterns that contain core uncovery time and vessel failure time were obtained for Young Gwang Nuclear (YGN) Units 3&4 using modular accident analysis program (MAAP) 3.0B code. Three different severe accident scenarios, such as two loss of coolant accidents (LOCAs) and station blackout(SBO), were considered in this analysis. Results indicated that design parameters related to refueling water storage tank (RWST), accumulator and steam generator (S/G) have more dominant effects on the progression of severe accidents investigated, compared to tile other six parameters.

• Journal of Communications and Networks
• /
• v.13 no.1
• /
• pp.1-5
• /
• 2011

Feedback with carry shift registers (FCSRs) over 2-adic number would be suitable in hardware implementation, but the are not efficient in software implementation since their basic unit (the size of register clls) is 1-bit. In order to improve the efficiency we consider FCSRs over $2^{\ell}$-adic number (i.e., FCSRs with register cells of size ${\ell}$-bit) that produce ${\ell}$ bits at every clocking where ${\ell}$ will be taken as the size of normal words in modern CPUs (e.g., ${\ell}$ = 32). But, it is difficult to deal with the carry that happens when the size of summation results exceeds that of normal words. We may use long variables (declared with 'unsigned _int64' or 'unsigned long long') or conditional operators (such as 'if' statement) to handle the carry, but both the arithmetic operators over long variables and the conditional operators are not efficient comparing with simple arithmetic operators (such as shifts, maskings, xors, modular additions, etc.) over variables of size ${\ell}$-hit. In this paper, we propose some conditions for FCSRs over $2^{\ell}$-adic number which admit fast software implementations using only simple operators. Moreover, we give two implementation examples for the FCSRs. Our simulation result shows that the proposed methods are twice more efficient than usual methods using conditional operators.