• ETRI Journal
• /
• 제44권4호
• /
• pp.695-707
• /
• 2022

As for the insufficient nature of the fossil fuel resources, the renewable energies as alternative fuels are imperative and highly heeded. To deliver the required electric power to the industrial and domestic consumers from DC renewable energy sources like fuel cell (FC), the power converter operates as an adjustable interface device. This paper suggests a new boost structure to provide the required voltage with wide range gain for FC power source. The proposed structure based on the boost converter and the quazi network, the so-called SBQN, can effectively enhance the FC functionality against its high operational sensitivity to experience low current ripple and also propagate voltage and current with low stress across its semiconductors. Furthermore, the switching power losses have been decreased to make this structure more durable. A full operational analysis of the proposed SBQN and its advantages over the conventional and famous structures has been compared and explained. Furthermore, a prototype of the single-phase converter has been constructed and tested in the laboratory.

• IMMUNE NETWORK
• /
• 제5권1호
• /
• pp.1-10
• /
• 2005

Background: Chronic infection with hepatitis B virus (HBV) affects about 350 million people worldwide, which have a high risk of development of cirrhosis and hepatocellular carcinoma. Treatment of chronic HBV infection relies on IFN-${\alpha}$ or lamivudine. However, interferon-${\alpha}$ is effective in only about 30% of patients. Also, the occurrence of escape mutations limits the usage of lamivudine. Therefore, the development and evaluation of new compounds or approaches are urgent. Methods: We comparatively evaluated DNA and adenoviral vaccines expressing HBV antigens, either alone or in combined regimens, for their ability to elicit Th1-type immune responses in Balb / c mice which are believed to be suited to resolve HBV infection. The vaccines were tested with or without a genetically engineered IL-12 (mIL-12 N220L) which was shown to enhance sustained Th1-type immune responses in HCV E2 DNA vaccine. Results: Considering the Th1-type cytokine secretion and the IgG2a titers, the strongest Th1-type immune response was elicited by the DNA prime-adenovirus boost regimen in the presence of mIL-12 N220L. In addition, the codelivery of mIL-12 N220L modulated differentially the immune responses by different vaccination regimens. Conclusion: Our results suggest that the DNA prime-adenovirus boost regimen in the presence of mIL-12 N220L may be the best candidate for HBV vaccine therapy of the regimens tested in this study and will be worthwhile being evaluated in chronic HBV patients.

• Journal of Electrical Engineering and Technology
• /
• 제12권3호
• /
• pp.1203-1210
• /
• 2017

Switched-inductor (SL) Z-source three-level inverter is a novel high power topology. The SL based impedance network can boost the input dc voltage to a higher value than the single LC impedance network. However, as all the neutral-point-clamped (NPC) inverters, the SL Z-source three-level inverter has to balance the neutral-point (NP) potential too. The principle of the inverter is introduced and then the effects of NP potential unbalance are analyzed. A NP balancing method is proposed. Other than the methods for conventional NPC inverter without Z-source impedance network, the upper and lower shoot-through durations are corrected by the feedforward compensation factors. With the proposed method, the NP potential is balanced and the voltage boosting ability of the Z-source network is not affected obviously. Simulations are conducted to verify the proposed method.

• 한국해양공학회지
• /
• 제34권6호
• /
• pp.461-468
• /
• 2020

This study examined how national research and development (R&D) in the domain of logistics has changed recently in the Republic of Korea. We conducted basic statistical analysis and social network analysis on 5,327 logistics-related R&D projects undertaken during 2005-2019. Data for performing these analyses were collected from the R&D database of the National Science and Technology Information Service (NTIS). By constructing a co-occurrence matrix with keywords, we conducted degree and betweenness centrality analysis and visualized the network matrix to display a cluster map. This study presents our observations related to the following findings: (1) the chronical trends of logistics R&D, (2) focused fields of logistics R&D, (3) the relations among keywords, and (4) the characteristics of logistics R&D. Finally, we suggest policy implications to boost and diversify logistics R&D.

• International Journal of Computer Science & Network Security
• /
• 제22권4호
• /
• pp.329-339
• /
• 2022

Small cells, particularly femtocells, are regarded a promising solution for limited resources required to handle the increasing data demand. They usually boost wireless network capacity. While widespread usage of femtocells increases network gain, it also raises several challenges. Interference is one of such concerns. Interference management is also seen as a main obstacle in the adoption of two-tier networks. For example, placing femtocells in a traditional macrocell's geographic area. Interference comes in two forms: cross-tier and co-tier. There have been previous studies conducted on the topic of interference management. This study investigates the principle of categorization of interference management systems. Many methods exist in the literature to reduce or eliminate the impacts of co-tier, cross-tier, or a combination of the two forms of interference. Following are some of the ways provided to manage interference: FFR, Cognitive Femtocell and Cooperative Resource Scheduling, Beamforming Strategy, Transmission Power Control, and Clustering/Graph-Based. Approaches, which were proposed to solve the interference problem, had been presented for each category in this work.

• 한국방사선학회논문지
• /
• 제17권3호
• /
• pp.423-431
• /
• 2023

본 연구의 목적은 환자의 신체정보와 인공지능 기법을 활용하여 부작용에 영향을 미치는 인자들을 분석하고 조영제 부작용의 정도를 예측하여 이를 완화하는 기초자료로 활용되고자 한다. 연구에 사용한 데이터는 서울 소재 종합병원의 검진을 시행한 CT 검사 58,000건 중 조영제 부작용이 발생한 1,235건 중 과거력 조사에서 조영제 부작용이 없었던 606명의 검사자를 대상자로 하였다. 606개 샘플 중 70%는 훈련 셋으로 사용하고 나머지 30%는 검증을 위한 테스트 셋으로 사용하였다. 나이, BMI(Body Mass Index), GFR(Glomerular Filtration Rate), BUN(Blood Urea Nitrogen), GGT(Gamma Glutamyl Transgerase), AST(Aspartate Amino Transferase,), and ALT(Alanine Amiono Transferase)의 feature를 독립변수로 조영제 중증도를 목표변수로 사용하였다. AdaBoost, Tree, Neural network, SVM, Random foest 알고리즘을 통해 AUC(Area under curve), CA(Classification Accuracy), F1, Precision, Recall을 파악하였다. 분류 예측에 사용된 알고리즘 중 가장 높은 평가지표를 나타내 것은 AdaBoost와 Random Forest이다. 모든 모델의 예측에서 가장 큰 요인은 GFR, BMI, GGT 이였다. 이는 신장 여과 기능, 비만에 따라 주입되는 조영제 양의 차이와 대사증후군의 여부에 따라 조영제 부작용 중증도에 영향을 미치는 것을 알 수 있었다.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 1998년도 Proceedings ICPE 98 1998 International Conference on Power Electronics
• /
• pp.966-971
• /
• 1998

The studies on the photovoltaic system are extensively exhaustible and broadly available resourse as a future energy supply. In this paper, a new maximum power point tracker(MPPT) using neural network theory is proposed to improve energy conversion efficiency. The boost converter and neural network controller(NNC) were employed so that the operating point of solar cell was located at the Maximum Power Point. And the back propagation algorithm with one input layer of two inputs(E, CE) and output layer(cnntrol value) was applied to train a neural network. Simulation and experimental results show that the performance of NNC in MPPT of photovoltaic array is better than that of controller based upon the Hill Climbing Method.

• Journal of Communications and Networks
• /
• 제10권1호
• /
• pp.89-97
• /
• 2008

Anomaly detection systems playa significant role in protection mechanism against attacks launched on a network. The greatest challenge in designing systems detecting anomalous exploits is defining what to measure. Effective yet simple, Shannon entropy metrics have been successfully used to detect specific types of malicious traffic in a number of commercially available IDS's. We believe that Renyi entropy measures can also adequately describe the characteristics of a network as a whole as well as detect abnormal traces in the observed traffic. In addition, Renyi entropy metrics might boost sensitivity of the methods when disambiguating certain anomalous patterns. In this paper we describe our efforts to understand how Renyi mutual information can be applied to anomaly detection as an offline computation. An initial analysis has been performed to determine how well fast spreading worms (Slammer, Code Red, and Welchia) can be detected using our technique. We use both synthetic and real data audits to illustrate the potentials of our method and provide a tentative explanation of the results.

• International Journal of Computer Science & Network Security
• /
• 제22권4호
• /
• pp.292-298
• /
• 2022

Femtocell networks can be a potential method for increasing the capacity of LTE networks, especially in indoor areas. However, unplanned deployment of femtocells results in co-tier interference and cross-tier interference problems. The interference reduces the advantages of implementing femtocell networks to a certain extent. The notion of Fractional Frequency Reuse (FFR) is proposed in order to reduce the impact of interference on the system's performance. In this paper, a dynamic approach for efficiently partitioning the spectrum is suggested. The goal is to enhance the capacity of femtocells, which will improve the performance of the system. The suggested strategy allocates less resources to the macrocell portion of the network, which has a greater number of femtocells deployed to maximize the utilization of available resources for femtocell users. The spectrum division would be dynamic. The proposed strategy is evaluated through a simulation using MATLAB tool. In conclusion, the results showed that the proposed scheme has the potential to boost the system's capacity.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2020년도 추계학술대회
• /
• pp.234-236
• /
• 2020

Content-adaptive training and transmission of the model parameters of neural networks can boost up the SR performance with higher restoration fidelity. In this case, efficient transmission of neural network parameters are essentially needed. Thus, we propose a novel method of compressing the network model parameters based on the training of network model parameters in the sense that the residues of filter parameters and content loss are jointly minimized. So, the residues of filter parameters are only transmitted to receiver sides for different temporal portions of video under consideration. This is advantage for image restoration applications with receivers (user terminals) of low complexity. In this case, the user terminals are assumed to have a limited computation and storage resource.