• Title/Summary/Keyword: energy space

Search Result 3,298, Processing Time 0.035 seconds

Toward Optimal FPGA Implementation of Deep Convolutional Neural Networks for Handwritten Hangul Character Recognition

  • Park, Hanwool;Yoo, Yechan;Park, Yoonjin;Lee, Changdae;Lee, Hakkyung;Kim, Injung;Yi, Kang
    • Journal of Computing Science and Engineering
    • /
    • v.12 no.1
    • /
    • pp.24-35
    • /
    • 2018
  • Deep convolutional neural network (DCNN) is an advanced technology in image recognition. Because of extreme computing resource requirements, DCNN implementation with software alone cannot achieve real-time requirement. Therefore, the need to implement DCNN accelerator hardware is increasing. In this paper, we present a field programmable gate array (FPGA)-based hardware accelerator design of DCNN targeting handwritten Hangul character recognition application. Also, we present design optimization techniques in SDAccel environments for searching the optimal FPGA design space. The techniques we used include memory access optimization and computing unit parallelism, and data conversion. We achieved about 11.19 ms recognition time per character with Xilinx FPGA accelerator. Our design optimization was performed with Xilinx HLS and SDAccel environment targeting Kintex XCKU115 FPGA from Xilinx. Our design outperforms CPU in terms of energy efficiency (the number of samples per unit energy) by 5.88 times, and GPGPU in terms of energy efficiency by 5 times. We expect the research results will be an alternative to GPGPU solution for real-time applications, especially in data centers or server farms where energy consumption is a critical problem.

A Study on Shear Strength under Constant Normal Load Conditions by Using 3DEC (3DEC을 이용한 일정수직하중 조건에서의 전단강도에 관한 연구)

  • Noh, Young-Mok;Mun, Hong-Ju;Kim, Ki-Ho;Jang, Won-Yil
    • Tunnel and Underground Space
    • /
    • v.24 no.1
    • /
    • pp.46-54
    • /
    • 2014
  • Direct shear tests have been initiated to understand the characteristics of joints which crucially affect the stability of rock mass. In this research, numerical approach in direct shear tests has been initiated using 3DEC on the basis of 3D distinct element method. Normal loads were altered in four different levels on artificial joint tests depending on the sawtooth angle and strengths on constant normal stress conditions, measuring the peak shear strength according to the direct shear tests under laboratory condition. Also results obtained from mechanical properties through laboratory test were used to perform numerical modeling, and shear strength obtained from the modeling was used to compare with laboratory direct shear test. As a result numerical analysis from distinct element method can simulate well on the shear behavior of rockmass.

Investigation of Frozen Rock Failure using Thermal Infrared Image (열적외선영상을 이용한 동결된 암석의 파괴특성 연구)

  • Park, Jihwan;Park, Hyeong-Dong
    • Tunnel and Underground Space
    • /
    • v.25 no.2
    • /
    • pp.144-154
    • /
    • 2015
  • Mechanical energy is accumulated in the object when stress is exerted on rock specimens, and the failure is occurred when the stress is larger than critical stress. The accumulated energy is emitted as various forms including physical deformation, light, heat and sound. Uniaxial compression strength test and point load strength test were carried out in low temperature environment, and thermal variation of rock specimens were observed and analyzed quantitatively using thermal infrared camera images. Temperature of failure plane was increased just before the failure because of concentration of stress, and was rapidly increased at the moment of the failure because of the emission of thermal energy. The variations of temperature were larger in diorite and basalt specimens which were strong and fresh than in tuff specimens which were weak and weathered. This study can be applied to prevent disasters in rock slope, tunnel and mine in cold regions and to analyze satellite image for predicting earthquake in cold regions.

Penetration of Compacted Bentonite into the Discontinuity in the Excavation Damaged Zone of Deposition Hole in the Geological Repository (심층처분장 처분공 주변 굴착손상영역에 존재하는 불연속면으로의 압축 벤토나이트 침투)

  • Lee, Changsoo;Cho, Won-Jin;Kim, Jin-Seop;Kim, Geon-Young
    • Tunnel and Underground Space
    • /
    • v.30 no.3
    • /
    • pp.193-213
    • /
    • 2020
  • A mathematical model to simulate more realistically the penetration of compacted bentonite buffer installed in the deposition hole into the discontinuity in the excavation damaged zone formed at the inner wall of the deposition hole in the geological repository for spent fuel is developed. In this model, the penetration of compacted bentonite is assumed to be the flow of Bingham fluid through the parallel planar rock fracture. The penetration of compacted bentonite is analyzed using the developed model. The results show that the maximum penetration depth of compacted bentonite into the rock fracture is proportioned to the swelling pressure of saturated compacted bentonite and the aperture of rock fracture. However, it is in inverse proportion to the yield strength of compacted bentonite. The viscosity of compacted bentonite dominates the penetration rate of compacted bentonite, but has no influence to the maximum penetration depth.

Solar Photovoltaics Technology: No longer an Outlier

  • Kazmerski, Lawrence L.
    • Proceedings of the Korean Vacuum Society Conference
    • /
    • 2011.08a
    • /
    • pp.70-70
    • /
    • 2011
  • The prospects of current and coming solar-photovoltaic (PV) technologies are envisioned, arguing this solar-electricity source is beyond a tipping point in the complex worldwide energy outlook. Truly, a revolution in both the technological advancements of solar PV and the deployment of this energy technology is underway; PV is no longer an outlier. The birth of modern photovoltaics (PV) traces only to the mid-1950s, with the Bell Telephone Laboratories' development of an efficient, single-crystal Si solar cell. Since then, Si has dominated the technology and the markets, from space through terrestrial applications. Recently, some significant shift toward technology diversity have taken place. Some focus of this presentation will be directed toward PV R&D and technology advances, with indications of the limitations and relative strengths of crystalline (Si and GaAs) and thin-film (a-Si:H, Si, Cu(In,Ga)(Se,S)2, CdTe). Recent advances, contributions, industry growth, and technological pathways for transformational now and near-term technologies (Si and primarily thin films) and status and forecasts for next-generation PV (nanotechnologies and non-conventional and "new-physics" approaches) are evaluated. The need for R&D accelerating the now and imminent (evolutionary) technologies balanced with work in mid-term (disruptive) approaches is highlighted. Moreover, technology progress and ownership for next generation solar PV mandates a balanced investment in research on longer-term (the revolution needs revolutionary approaches to sustain itself) technologies (quantum dots, multi-multijunctions, intermediate-band concepts, nanotubes, bio-inspired, thermophotonics, ${\ldots}$ and solar hydrogen) having high-risk, but extremely high performance and cost returns for our next generations of energy consumers. This presentation provides insights to the reasons for PV technology emergence, how these technologies have to be developed (an appreciation of the history of solar PV)-and where we can expect to be by this mid-21st century.

  • PDF

A study on thermal simulation for extensive green roof system using a plant canopy model (식생캐노피모델을 통한 저관리 조방형 옥상녹화시스템의 열해석 전산모의에 관한 연구)

  • Kim, Tae Han
    • Journal of the Korean Society of Environmental Restoration Technology
    • /
    • v.15 no.2
    • /
    • pp.137-147
    • /
    • 2012
  • GRS is an effective urban ecology restoration technique that can manage a variety of environmental functions such as ecological restoration, rainwater spill control and island heat effect from a low-impact development standpoint that can be utilized in new construction and retrofits. Recently, quantitative evaluation studies, both domestic and abroad, in the areas related to these functions, including near-earth surface climate phenomenon, heavy rainwater regulation, thermal environment of buildings, have been actively underway, and there is a trend to standardize in the form of technological standards. In particular, centered on the advanced European countries, studies of standardizing the specific insulation capability of buildings with green system that comprehensively includes the green roof, from the perspective of replacing the exterior materials of existing buildings, are in progress. The limitation of related studies in the difficulties associated with deriving results that reflect material characteristics of continuously evolving systems due in part to not having sufficiently considered the main components of green system, mechanisms of vegetation, soils. This study attempts to derive, through EnergyPlus, the effects that the vegetation-related indicators such as vegetation height, FCV, etc. have on building energy load, by interpreting vegetation and soil mechanisms through plant canopy model and using an ecological standard indicator LAI that represent the condition of plant growth. Through this, the interpretations that assume green roof system as simple heat insulation will be complemented and a more practical building energy performance evaluation method that reflects numerical methods for heat fluxes phenomena that occur between ecology restoration systems comprised of plants and soil and the ambient space.

Temperature Dependent Cation Distribution in Tb2Bi1Ga1Fe4O12

  • Park, Il-Jin;Park, Chu-Sik;Kang, Kyoung-Soo;Kim, Chul-Sung
    • Journal of Magnetics
    • /
    • v.13 no.3
    • /
    • pp.110-113
    • /
    • 2008
  • In this study, heavy rare earth garnet $Tb_2Bi_1Ga_1Fe_4O_{12}$ powders were fabricated by a sol-gel and vacuum annealing process. The crystal structure was found to be single-phase garnet with a space group of Ia3d. The lattice constant $a_0$ was determined to be 12.465 ${\AA}$. From the analysis of the vibrating sample magnetometer (VSM) hysteresis loop at room temperature, the saturation magnetization and coercivity of the sample are 7.64 emu/g and 229 Oe, respectively. The N$\acute{e}$el temperature($T_N$) was determined to be 525 K. The M$\ddot{o}$ssbauer spectrum of $Tb_2Bi_1Ga_1Fe_4O_{12}$ at room temperature consists of 2 sets of 6 Lorentzians, which is the pattern of single-phase garnet. From the results of the M$\ddot{o}$ssbauer spectrum at room temperature, the absorption area ratios of Fe ions on 24d and 16a sites are 74.7% and 25.3%(approximately 3:1), respectively. These results show that all of the non-magnetic Ga atoms occupy the 16a site by a vacuum annealing process. Absorption area ratios of Fe ions are dependent not only on a sintering condition but also on the temperature of the sample. It can then be interpreted that the Ga ion distribution is dependent on the temperature of the sample. The M$\ddot{o}$ssbauer measurement was carried out in order to investigate the atomic migration in $Tb_2Bi_1Ga_1Fe_4O_{12}$.

Facial Feature Extraction Using Energy Probability in Frequency Domain (주파수 영역에서 에너지 확률을 이용한 얼굴 특징 추출)

  • Choi Jean;Chung Yns-Su;Kim Ki-Hyun;Yoo Jang-Hee
    • Journal of the Institute of Electronics Engineers of Korea SP
    • /
    • v.43 no.4 s.310
    • /
    • pp.87-95
    • /
    • 2006
  • In this paper, we propose a novel feature extraction method for face recognition, based on Discrete Cosine Transform (DCT), Energy Probability (EP), and Linear Discriminant Analysis (LDA). We define an energy probability as magnitude of effective information and it is used to create a frequency mask in OCT domain. The feature extraction method consists of three steps; i) the spatial domain of face images is transformed into the frequency domain called OCT domain; ii) energy property is applied on DCT domain that acquire from face image for the purpose of dimension reduction of data and optimization of valid information; iii) in order to obtain the most significant and invariant feature of face images, LDA is applied to the data extracted using frequency mask. In experiments, the recognition rate is 96.8% in ETRI database and 100% in ORL database. The proposed method has been shown improvements on the dimension reduction of feature space and the face recognition over the previously proposed methods.

Appilication of a Green City Index as a Green Space Planning Index for the Low-Carbon Green City of Gangneung-si

  • Cho, Su-Hyun;Jo, Hyun-Ju
    • Journal of Environmental Science International
    • /
    • v.25 no.10
    • /
    • pp.1381-1387
    • /
    • 2016
  • This study aims to establish baseline data for sustainable monitoring by applying the green city index (GCI), which is set up to evaluate the city level, to the city of Gangneung-si, which was designated as a pilot city for the Low-carbon Green Growth City project by the Ministry of Land, Infrastructure, and Transportation. The GCI was applied in the framework of European systems, while considering the social and economic status of Korea. Indicators from 7 areas-$CO_2$, energy, building, transportation, water, waste, and quality of atmosphere were analyzed, except for qualitative indicators. Results indicate that total $CO_2$ emissions were 30.8 tons per capita, or 2.2 tons per one million units of real GDP. The total final energy consumption was 0.231 TOE/capita, or 0.317 TOE per one million units of real GDP. The percentage of total energy derived from renewable resources was 0.41% and energy consumption by the building was $433.5Mwh/1,000m^2$. The total percentage of the working population travelling to work daily by public transportation (limited to bus) was 19%. Further, the total annual water consumption was $99m^3/capita$, and the water lost in the water distribution system was $0.057m^3/capita/day$. The total annual waste collected was 0.0077 ton per capita, The annual mean emission were 0.014 ppm/day for $NO_2$, 0.005 ppm/day for $SO_2$, and 0.019 ppm/day for $O_3$. The annual mean for PM10 emissions was $39{\mu}g/m^3/day$.

The Study of CsI(Tl) Scintillation Detector Design and Signal Processing for the Measurement of the Radiation Distribution (방사선 분포측정용 CsI(Tl) 검출기 설계 및 신호처리에 관한 연구)

  • Hwang, Young-gwan;Lee, Nam-ho;Kim, Jong-yeol;Jeong, Sang-hun
    • Proceedings of the Korean Institute of Information and Commucation Sciences Conference
    • /
    • 2016.05a
    • /
    • pp.778-779
    • /
    • 2016
  • In This paper, We designed the scintillation detector for measuring radiation signals in units of pixels for a radiation source that is distributed in the space. And we carried out a study to design a radiation imaging by the module for obtaining the detection signal. For measuring radiation distribution we configure a radiation detector combining CsI(Tl) scintillator and a photodiode. In addition, its performance was verified via gamma irradiation test.

  • PDF