都市地價 構成要因의 價格推定에 關한 時系列的 硏究 : 헤도닉 가격함수<hedonic price function>를 中心으로

이명호 홍익대학교 1999 국내석사

고저로 대역에서 마이크폰 어레이를 사용한 음성입력에 관한 연구

이명호 숭실대학교 1998 국내석사

Task mapping algorithms for embedded signal processing

이명호 University of Southern California 1999 해외박사

In this dissertation, we develop task mapping algorithms for Embedded Signal Processing (ESP) applications such as radar and sonar signal processing. Computationally demanding algorithms have been recently proposed for these applications which provide improved performance for target detection. These applications are typically composed of a sequence of computation stages with varying computational complexity. Each stage consists of a number of identical tasks. Hard real-time performance constraints such as latency (i.e., the time for processing one input data set) and/or throughput (i.e., the number of outputs per unit time) are to be met in these applications. In order to achieve high performance for these applications, efficient task mapping is needed. In this dissertation, our main focus is on task mapping for optimizing throughput performance. Traditionally, scientific and engineering applications, which typically perform large-scale simulations, have not paid much attention to optimizing throughput performance. However, in ESP applications, a stream of input data is received. The output must also be produced continuously to keep pace with the data input rate. Therefore, throughput is a key performance measure in these applications. In this dissertation, we develop an algorithmic framework and techniques for task mapping to optimize throughput performance. We first address the algorithmic framework for task mapping. A task model is defined for ESP applications by considering their computational characteristics. Also the characteristics of the target machine platforms-message-passing High Performance Computing (HPC) platforms-are considered in defining the task model. We then discuss the execution model which is used for mapping tasks of ESP applications onto HPC platforms. We also describe algorithmic techniques such as data remapping between successive computation stages, clustering adjacent computation stages, and replication of computation stages. Our framework is then illustrated based on the task model, the execution model, and the algorithmic techniques. We propose two task mapping algorithms for optimizing throughput performance. These algorithms are based on a new execution model which relaxes the restricted mapping choices allowed in the previously used linear execution model. A novel stage partitioning technique is used to realize the new execution model. Based on the new model, we first develop a task mapping algorithm using dynamic programming solution. The time complexity using the dynamic programming solution, however, is high. Thus, we further develop a heuristic-based fast task mapping algorithm. This results in a fast task mapping, yet provides more efficient task mapping than previous approaches using the linear execution model. We implemented an example Space-Time Adaptive Processing (STAP) benchmark using the heuristic-based task mapping algorithm to optimize the throughput performance. Our code is written using C, LAPACK (a standard linear algebra library), and the Message Passing Interface (MPI). Therefore, our code can be easily ported to various HPC platforms. Experimental results on IBM SP2 and SGI/Cray T3E show the effectiveness of our approach compared with the previous approaches.

CaO-SiO_(2)-CaF_(2)계 슬래그의 질소용해반응에 따른 열역학적 연구

이명호 충남대학교 산업대학원 2002 국내석사

The nitrogen solubility and nitride capacity of CaO-SiO₂-CaF₂system were measured, using a gas-liquid equilibration technique. This system was chosen because it has wide liquidus region at the temperature of interest and was expected to show two types of nitrogen dissolution mechanism. The effect of oxygen partial pressure on the nitrogen solubility was examined the effects by varying the CO pressure between 0.6 and 0.3 atm with nitrogen pressure fixed at 0.4 atm. The linear relationship between nitrogen contents and oxygen partial pressure is clearly visible, with close to the theoretical value of -3/4. The effect of temperature on the nitride capacity of the salg was studied in the temperature range of 1350 to 1500℃. As the temperature increased, the nitrogen solubility increased, which inferred that nitrogen dissolution in the slag is an endothermic reaction. The values of the nitride capacity has a minimum at about 2.0 of basicity. This may be explained by two mechanisms for nitrogen dissolution in the slag; "incorporated nitride ion" and "free nitride ion" mechanism. These complex relationship between basicity of slag and nitride capacity is explained by using optical basicity also. It was found that nitride capacity and optical basicity had a clear relationship as following equations. logC_N=3.14A - 12.65(0.66≤A<0.77) logC_N=2.8A - 12.77(0.77≤A<0.84)

繁殖障害牛에 대한 鍼治療의 效果에 관한 硏究

이명호 建國大學校 農畜大學院 1995 국내석사

Eighty-seven cows having reproductive disorders such as repeat breeder (11 cows), anestrum (39 cows), Weak estrum (22 cows) and follicular cyst (15 cows) were treated by one-point acupuncture at Kyo-So Hyul from August of 1933 to August of 1994. The results from the study are as follows. 1. Eight(72.7%) of 11 cows with repeat breeder shows therapeutic effects after treatment of acupuncture, and 6 of them were pregnant. However, 3 cases were considered as non-effective because no significant clinical improvement were observed. 2. Thirty-one(79.5%) of 39 cows with anestrum symptoms were apparently improved. Twenty-six cases of them were pregnant, however, the remaint were not-pregnnt. 3. Fifteen(68.2%) of 22 cows having weak estrum showed therapeutic effects. Seven of them were not considered as effective because they were relatively young. 4. The follicular cyst of 15 cows were luteinized and atrophied after acupuncture treatment. The cows showing therpeutic effects were only 4 cases(26.7%)

住. 商複合 建築物의 設計 方法에 관한 硏究

이명호 嶺南大學校 1994 국내석사

흄 管製造를 위한 最適作業條件에 關한 實驗的 硏究

이명호 建國大學校 1980 국내석사

觀光호텔 廚房의 基本運營에 관한 硏究

이명호 京畿大學校 1990 국내석사

LNG냉열이용 부유식 냉동 및 발전 복합플랜트 설계에 관한 연구

이명호 한국해양대학교 대학원 2017 국내박사

Base on the energy diversification policy, Korea Gas Corporation started to import LNG from Indonesia in October 1986, and supplied natural gas to Gyeongin area through commercial operation of Pyeongtaek LNG terminal and imported from Malaysia in September 1991 also from July 1993 began to supplying to the Central area and also to the Yeongnam and Honam area. After that, commercial operation was started at the Incheon LNG terminal and the gas was supplied to the Busan area. In the first decade since importing, LNG annual use amount is 17 million tons, the amount used every year increases, the basic fuel for catering and heating necessary for daily routine for the first time in 20 years has changed from briquettes and oil to city gas. In order to reduce the volume of NG, liquefy the temperature to -162 ° C, transport it to the LNG carrier, store it in the LNG storage tank of Incheon, Pyeongtaek, Tongyoung and Gwangyang terminal and then supply to consumers after revaporizing it. Methane, which is the main component of natural gas used as city gas, supplies high energy of 13,000 kcal/kg in the combustion process, so it plays a role as an excellent energy source, and to supply it to customers, LNG using a seawater type carburetor to heat to normal temperature NG, 200 kcal/kg of cold heat which is 2.5 times higher than the heat of ice melting which is the core heat source of ice heat storage is discarded to the sea. The theoretical method for recovering the LNG cold which is discarded in this manner can use -155℃ cold heat generated in the vaporization process of LNG for liquefaction separation of air, released after liquefaction separation -130'C cold heat used for low temperature crushing of waste tires -100'C cold heat is used for metal forming and freezing warehouse and ice making and the cold heat generated after ice making can be supplied to cold heat required for cooling. For this reason, academia and related companies have made great efforts, but in Korea, it is possible to use cold energy only if it is adjacent to the piping system supplied from the LNG receiving terminal. Due to the geopolitical restriction, it is judged that it can not recover the huge amount of cold heat released in the process of converting to gas and it is not economical. However, in this study, we will focus on not only solving the above problems but also being able to recover enormous LNG cold heat if it is adjacent to the LNG supplier and the utilization recipient of LNG cold energy, I decided to do research on a freezing warehouse using LNG cold energy. In the neighboring areas of the island of Indonesia, which is a producer and exporter of LNG, the existing mechanical compression refrigeration system is used for freezing fisheries such as tuna, which is a high quality fish family. In addition, LNG vaporization system is installed and NG is supplied to power production facilities reaching 100MW. Therefore, a new concept, floating type Combined Cold storage and Power generator(CCP). This supply quick freezing warehouse using LNG cold energy and electricity from power generators using vaporized NG installed in the barge to Southeast Asia including Indonesia.

α-Tricalcium phosphate 와 혈소판농축혈장의 조합처리가 골모세포의 활성에 미치는 영향

이명호 원광대학교 2004 국내박사