• Journal of Korea Water Resources Association
• /
• v.57 no.3
• /
• pp.151-164
• /
• 2024

In water treatment plants supplying potable water, the management of chlorine concentration in water treatment processes involving pre-chlorination or intermediate chlorination requires process control. To address this, research has been conducted on water quality prediction techniques utilizing AI technology. This study developed an AI-based predictive model for automating the process control of chlorine disinfection, targeting the prediction of residual chlorine concentration downstream of sedimentation basins in water treatment processes. The AI-based model, which learns from past water quality observation data to predict future water quality, offers a simpler and more efficient approach compared to complex physicochemical and biological water quality models. The model was tested by predicting the residual chlorine concentration downstream of the sedimentation basins at Plant, using multiple regression models and AI-based models like Random Forest and LSTM, and the results were compared. For optimal prediction of residual chlorine concentration, the input-output structure of the AI model included the residual chlorine concentration upstream of the sedimentation basin, turbidity, pH, water temperature, electrical conductivity, inflow of raw water, alkalinity, NH₃, etc. as independent variables, and the desired residual chlorine concentration of the effluent from the sedimentation basin as the dependent variable. The independent variables were selected from observable data at the water treatment plant, which are influential on the residual chlorine concentration downstream of the sedimentation basin. The analysis showed that, for Plant, the model based on Random Forest had the lowest error compared to multiple regression models, neural network models, model trees, and other Random Forest models. The optimal predicted residual chlorine concentration downstream of the sedimentation basin presented in this study is expected to enable real-time control of chlorine dosing in previous treatment stages, thereby enhancing water treatment efficiency and reducing chemical costs.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.21 no.1
• /
• pp.29-41
• /
• 2019

Terrestrial Gross Primary Production (GPP) is the largest global carbon flux, and forest ecosystems are important because of the ability to store much more significant amounts of carbon than other terrestrial ecosystems. There have been several attempts to estimate GPP using mechanism-based models. However, mechanism-based models including biological, chemical, and physical processes are limited due to a lack of flexibility in predicting non-stationary ecological processes, which are caused by a local and global change. Instead mechanism-free methods are strongly recommended to estimate nonlinear dynamics that occur in nature like GPP. Therefore, we used the mechanism-free machine learning techniques to estimate the daily GPP. In this study, support vector machine (SVM), random forest (RF) and artificial neural network (ANN) were used and compared with the traditional multiple linear regression model (LM). MODIS products and meteorological parameters from eddy covariance data were employed to train the machine learning and LM models from 2006 to 2013. GPP prediction models were compared with daily GPP from eddy covariance measurement in a deciduous forest in South Korea in 2014 and 2015. Statistical analysis including correlation coefficient (R), root mean square error (RMSE) and mean squared error (MSE) were used to evaluate the performance of models. In general, the models from machine-learning algorithms (R = 0.85 - 0.93, MSE = 1.00 - 2.05, p < 0.001) showed better performance than linear regression model (R = 0.82 - 0.92, MSE = 1.24 - 2.45, p < 0.001). These results provide insight into high predictability and the possibility of expansion through the use of the mechanism-free machine-learning models and remote sensing for predicting non-stationary ecological processes such as seasonal GPP.

• Clean Technology
• /
• v.26 no.2
• /
• pp.102-108
• /
• 2020

In recent years, packed column has been widely used in separation processes, such as absorption, desorption, distillation, and extraction, in the petrochemical, fine chemistry, and environmental industries. Packed column is used as a contacting facility for gas-liquid and liquid-liquid systems filled with random packed materials in the column. Packed column has various advantages such as low pressure drop, economical efficiency, thermally sensitive liquids, easy repairing restoration, and noxious gas treatment. The performance of a packed column is highly dependent on the maintenance of good gas and liquid distribution throughout a packed bed; thus, this is an important consideration in a design of packed column. In this study, hydraulic pressure drop, hold-up as a function of liquid load, and mass transfer in the air, air/water, and air-NH₃/water systems were studied to find the geometrical characteristic for raschig super-ring experiment dry pressure drop. Based on the results, design factors and operating conditions to handle noxious gases were obtained. The dry pressure drop of the random packing raschig super-ring was linearly increased as a function of gas capacity factor with various liquid loads in the Air/Water system. This result is lower than that of 35 mm Pall-ring, which is most commonly used in the industrial field. Also, it can be found that the hydraulic pressure drop of raschig super-ring is consistently increased by gas capacity factor with various liquid loads. When gas capacity factor with various liquid loads is increased from 1.855 to 2.323 kg^-1/2 m^-1/2 S^-1, hydraulic pressure drop increases around 17%. Finally, the liquid hold-up related to packing volume, which is a parameter of specific liquid load depending on gas capacity factor, shows consistent increase by around 3.84 kg^-1/2 m^-1/2 S^-1 of the gas capacity factor. However, liquid hold-up significantly increases above it.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.11 no.4
• /
• pp.281-291
• /
• 2013

A hypothetical repository was assumed to be located at the KURT (KAERI Underground Research Tunnel) site, and the travel times of radionuclides released from three source positions were calculated. The groundwater flow around the KURT site was simulated and the groundwater pathways from the hypothetical source positions to the shallow groundwater were identified. Of the pathways, three pathways were selected because they had highly water-conductive features. The transport travel times of the radionuclides were calculated by a TDRW (Time-Domain Random Walk) method. Diffusion and sorption mechanisms in a host rock matrix as well as advection-dispersion mechanisms under the KURT field condition were considered. To reflect the radioactive decay, four decay chains with the radionuclides included in the high-level radioactive wastes were selected. From the simulation results, the half-life and distribution coefficient in the rock matrix, as well as multiple pathways, had an influence on the mass flux of the radionuclides. For enhancing the reliability of safety assessment, this reveals that identifying the history of the radionuclides contained in the high-level wastes and investigating the sorption processes between the radionuclides and the rock matrix in the field condition are preferentially necessary.

• Journal of KIISE:Information Networking
• /
• v.31 no.5
• /
• pp.490-500
• /
• 2004

In mobile/wireless environment, mobility management is widely being focused as one popular researches. But, disruption happens when messages are exchanged between nodes as registration is made after handoff, and unnecessary traffic occurs because of the use of the Random-walk model, in which the probability for MN to move to neighboring cells is equal. In order to solve these problems, this study proposes a technique and algorithm for composing Directional Shadow Registration Region (DSRR) that provides seamless mobility. The core of DSRR is to prevent disruption and unnecessary traffic by minimizing the number o) neighboring cells with a high probability of handoff (AAAF). This study sensed the optimal time for handoff through regional cell division by introducing a division scheme, and then decided DSRR, the region for shadow registration, by applying direction vector (DV) obtained through directional cell sectoring. According to the result of the experiment, the proposed DSRR processes message exchange between nodes within the intra-domain, the frequency of disruptions decreased significantly compared to that in previous researches that process in inter-domain environment. In addition, traffic that occurs at every handoff happened twice in DSRR compared to n (the number of neighboring cells) times in Previous researches. As an additional effect, divided regions obtained from the process of composing DSRR filter MN that moves regardless of handoff.

• Journal of IKEEE
• /
• v.21 no.2
• /
• pp.136-141
• /
• 2017

In this paper, we propose an algorithm for memory optimization to apply deep learning to PC. The proposed algorithm minimizes the memory and computation processing time by reducing the amount of computation processing and data required in the conventional deep learning structure in a general PC. The algorithm proposed in this paper consists of three steps: a convolution layer configuration process using a random filter with discriminating power, a data reduction process using PCA, and a CNN structure creation using SVM. The learning process is not necessary in the convolution layer construction process using the discriminating random filter, thereby shortening the learning time of the overall deep learning. PCA reduces the amount of memory and computation throughput. The creation of the CNN structure using SVM maximizes the effect of reducing the amount of memory and computational throughput required. In order to evaluate the performance of the proposed algorithm, we experimented with Yale University's Extended Yale B face database. The results show that the algorithm proposed in this paper has a similar performance recognition rate compared with the existing CNN algorithm. And it was confirmed to be excellent. Based on the algorithm proposed in this paper, it is expected that a deep learning algorithm with many data and computation processes can be implemented in a general PC.

• Journal of Microbiology and Biotechnology
• /
• v.28 no.11
• /
• pp.1916-1927
• /
• 2018

Corynebacterium glutamicum is an excellent platform for the production of amino acids, and is widely used in the fermentation industry. Most industrial strains are traditionally obtained by repeated processes of random mutation and selection, but the genotype of these strains is often unclear owing to the absence of genomic information. As such, it is difficult to improve the growth and amino acid production of these strains via metabolic engineering. In this study, we generated a complete genome map of an industrial L-valine-producing strain, C. glutamicum XV. In order to establish the relationship between genotypes and physiological characteristics, a comparative genomic analysis was performed to explore the core genome, structural variations, and gene mutations referring to an industrial L-leucine-producing strain, C. glutamicum CP, and the widely used C. glutamicum ATCC 13032. The results indicate that a 36,349 bp repeat sequence in the CP genome contained an additional copy each of lrp and brnFE genes, which benefited the export of L-leucine. However, in XV, the kgd and panB genes were disrupted by nucleotide insertion, which increase the availability of precursors to synthesize L-valine. Moreover, the specific amino acid substitutions in key enzymes increased their activities. Additionally, a novel strategy is proposed to remodel central carbon metabolism and reduce pyruvate consumption without having a negative impact on cell growth by introducing the CP-derived mutant $H^+$/citrate symporter. These results further our understanding regarding the metabolic networks in these strains and help to elucidate the influence of different genotypes on these processes.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.35 no.8
• /
• pp.1088-1096
• /
• 2006

The purpose of this study was to identify the service encounter blueprint by types of restaurants in order to manage moment of truth when customers who visit a restaurant encounter services. The service encounter blueprint gives an overall picture of the service provision to visualize an entire service process and its integrated structure. The blueprint is used for service process analysis technique. The random samples of 15 customers were observed by types of restaurants and the records were collected for three-days' observation. Interviews were performed by 3 managers, 3 service encounter employees, 3 cashiers, 3 cooks and 10 customers by types of restaurants. After drawing the first service blueprint, it was revised by the interview with the 3 managers and 6 service encounter employees. In this paper, restaurant service processes are reviewed and analyzed. By use of service blueprint, the processes are analyzed to find a fail point, customer wait, employee decision. As a result of making a blueprint of service encounter by types of restaurant, blueprints of fine-dining restaurants and family restaurants were similar, while fast-food restaurants showed a little difference. In particular, difference was indicated in a point where interaction of service encounter occurred. Difference was indicated depending on types of restaurants. Therefore, the efforts to improve this problem were needed. The blueprint is a map or flowchart (called a process chart in manufacturing) of all transactions constituting the service delivery process. The results showed that service encounter blueprint can be used to improve the service process in the restaurant's encounter.

• Proceedings of the Korean Society of Developmental Biology Conference
• /
• 2003.10a
• /
• pp.84-84
• /
• 2003

DNA methylation is involved in epigenetic processes such as X-chromosome inactivation, imprinting and silencing of transposons. DNA methylation is a highly plastic and critical component of mammalian development The DNA methyltransferases (Dnmts) are responsible for the generation of genomic methylation patterns, which lead to transcriptional silencing. The maintenance DNA methyltransferase enzyme, Dnmt 1, and the de novo methyltransferase, Dnmt3a and Dnmt3b, are indispensable for development because mice homozygous for the targeted disruption of any of these genes are not viable. The occurrence of DNA methylation is not random, and it can result in gene silencing The mechanisms underlying these processes are poorly understood. It is well established that DNA methylation and histone deacetylation operate along a common mechanistic pathway to repress transcription through the action of methyl-binding domain proteins (MBDs), which are components of, or recruit, histone deacetylase (HDAC) complexes to methylated DNA. As a basis for future studies on the role of the DNA-methyl-transferase in porcine development, we have isolated and characterized a partial cDNA coding for the porcine Dnmt1. Total RNA of testis, lung and ovary was isolated with TRlzol according to the manufacture's specifications. 5 ug of total RNA was reverse transcribed with Super Script II in the presence of porcine Dnmt 1 specific primers. Standard PCRs were performed in a total volume of 50 ul with cDNA as template. Two DNA fragmenets in different position were produced about 700bp, 1500bp and were cloned into pCR II-TOPO according to the manufacture's specification. Assembly of all sequences resulted in a cDNA from 158bp of 5'to 4861bp of 3'compare with the known human maintenance methyltransferase. Now, we are cloning the unknown Dnmt 1 region by 5'-RACE method and expression of Dnmt 1 in tissues from adult porcine animals.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.14 no.4
• /
• pp.308-318
• /
• 2002

The requirements to the numerical model of wind-generated waves in shallow water are discussed in the framework of the Korteweg-de Vries equation. The weakness of nonlinearity and dispersion required for the Korteweg-de Vries equation applicability is considered for fully developed sea, non-stationary wind waves and swell, including some experimental data. We note for sufficient evaluation of the freak wave statistics it is necessary to consider more than about 10,000 waves in the wave record, and this leads to the limitation of the numerical domain and number of realizations. The numerical modelling of irregular water waves is made to demonstrate the possibility of effective evaluation of the statistical properties of freak waves with heights equal to 2-2.3 significant wave height.