• Title/Summary/Keyword: 인공환경

Search Result 3,542, Processing Time 0.032 seconds

Simulation of wetland storage volume using a classification-based artificial intelligence prediction model (분류 기반의 인공지능 예측 모델을 이용한 습지 저류량 모의)

  • Ji yu Seo;Ha eun Jung;Jeong Hoon Lee;Sang Dan Kim
    • Proceedings of the Korea Water Resources Association Conference
    • /
    • 2023.05a
    • /
    • pp.270-270
    • /
    • 2023
  • 습지 생태계는 복잡한 물리적 생지화학적 프로세스의 상호작용이 있으나, 습지 생태계의 건강성 회복을 위한 첫 번째 단계는 습지 생태계에서의 물순환에 대한 정확한 이해일 것이다. 또한 지역적인 물 균형 및 생태계 보전에서 습지를 활용하기 위해서도 습지 물순환에 대한 정량적인 이해는 필수적이다. 그러나, 습지 물순환의 이해를 위해 필수적인 관측 자료들은 현장 측정으로 획득하기 어려운 자료이거나 비용적인 문제로 인하여 관측이 어려운 실정이다. 이에 본 연구에서는 Sentinel-2 위성 자료를 활용하여 습지의 유입량을 추정하기 위한 절차를 제시하고자 한다. 이를 위해 한반도 동남부의 낙동강에 위치한 주요 다목적댐의 자료를 활용한 분류 기반의 인공지능 모델이 설계된다. 인공지능의 학습을 위한 입력자료는 아래와 같은 절차에 의해 만들어진다. 1) 다목적댐의 수위-물 체적 관계를 이용하여 수위-수표면적 관계 곡선 도출. 2) 수위-수표면적 관계 곡선과 DEM을 활용하여 물과 육지 영역을 구분하는 식별자를 도출. 3) Sentinel-2 위성 정보와 물-육지 식별자를 비교하는 랜덤 포레스트 모델을 설계. 4) 위성 정보의 물-육지 정보로부터 미계측 습지 지역의 물과 육지를 식별할 수 있는 식별자 도출. 이러한 과정을 경유하여 추정된 습지의 수표면적과 습지 지역의 DEM을 결합함으로써 습지의 수위-수표면적-물 체적 관계 곡선이 산정되어, 최종적으로 습지의 유입량이 모의된다. 모의된 습지 유입량은 다양한 수문 모델의 매개변수를 추정하는데 활용될 수 있을 것이며, 검증된 수문 모델을 활용하여 습지의 물순환의 이해도를 증진시킬 수 있을 것으로 기대된다.

  • PDF

A Monitoring Scheme Based on Artificial Intelligence in Mobile Edge Cloud Computing Environments (모바일 엣지 클라우드 환경에서 인공지능 기반 모니터링 기법)

  • Lim, JongBeom;Choi, HeeSeok;Yu, HeonChang
    • KIPS Transactions on Computer and Communication Systems
    • /
    • v.7 no.2
    • /
    • pp.27-32
    • /
    • 2018
  • One of the crucial issues in mobile edge cloud computing environments is to monitor mobile devices. Due to the inherit properties of mobile devices, they are prone to unstable behavior that leads to failures. In order to satisfy the service level agreement (SLA), the mobile edge cloud administrators should take appropriate measures through a monitoring scheme. In this paper, we propose a monitoring scheme of mobile devices based on artificial intelligence in mobile edge cloud computing environments. The proposed monitoring scheme is able to measure faults of mobile devices based on previous and current monitoring information. To this end, we adapt the hidden markov chain model, one of the artificial intelligence technologies, to monitor mobile devices. We validate our monitoring scheme based on the hidden markov chain model. The proposed monitoring scheme can also be used in general cloud computing environments to monitor virtual machines.

Assessment of Blocking Effect of Natural and Artificial Topography on Sunshine Duration Using GIS Data and Sunshine Model (GIS 자료와 일조모델을 이용한 자연적 및 인공적 지형에 의한 일조차단 평가)

  • Kim, Do Yong;Kim, Jae Jin
    • Journal of Korean Society for Geospatial Information Science
    • /
    • v.24 no.3
    • /
    • pp.67-73
    • /
    • 2016
  • The present study evaluated the blocking effect of natural and artificial topography on sunshine duration in the southern coastal area of Haui-do. The geospatial data for the target area was constructed by geographic information system(GIS) data. Three-dimensional modeling based on solar azimuth and altitude angles was conducted for the assessment of sunshine environment. The sunshine area was evaluated over 80~90% of the target area in the daytime, especially in summer. The blocking effect of mountainous terrain on sunshine duration was presented at the northern residential area in the late afternoon. There was also the effect of artificial topography by construction of fill-up bank on sunshine environment at the southern residential area early in the morning and the south-western part of salt field in the late afternoon.

Study on the Growth of Soybean and Corn in Artificial Media (인공배지를 이용한 옥수수와 콩의 생육시험 연구)

  • Kim, Sun-Joo;Yoon, Chun-Gyeong;Kim, Hyung-Joong;Yang, Yong-Suck
    • Korean Journal of Environmental Agriculture
    • /
    • v.19 no.3
    • /
    • pp.234-241
    • /
    • 2000
  • For the recycling of sludge as soil conditioner, the firing technology in pottery industry was applied to the sludge treatment, and produced artificial media with many voids. To produce artificial media using sludge, chabazite and lime were used as additives, and the mixture of sludge and additives was thermally treated in the firing kiln at about $800{\sim}1,100^{\circ}C$ for about ten minutes. The effect of mixed artificial media into upland soil was investigated through the crop growth experiment and the physical & chemical characteristics of the mixed soils were analysed. The pH of artificial media was higher than that of the control soils. After the plant growth experiment, artificial media plots almost contained more CEC, OM, TN, TP and $AV.P_2O_5$ than upland soil plots. From the growth analysis, growth of soybean and corn in the artificial media plots was better than that in the original upland soil plots. The yield of soybean and corn in the artificial media plots were about 46kg/10a, 194kg/10a, respectively, which is higher than that in the control. Heavy metals in the artificial media plots were lower than the standard regulation. Therefore, the artificial media produced from sludge can be mixed into farmland, and crop production can be increased additionally.

  • PDF

A Comparative Study on Music Preference between School-Aged Children with Cochlear Implant and Normal Hearing (학령기 인공와우 이식 아동과 일반 건청 아동의 음악 선호도에 대한 비교 연구)

  • Hwang, Sun Young
    • Journal of Music and Human Behavior
    • /
    • v.8 no.2
    • /
    • pp.47-64
    • /
    • 2011
  • This study was to examine music preferences of school-aged cochlear implant(CI) recipients in comparison with normal hearing(NH) children. Participants were 12 CI recipients and 155 NH children. CI children were recruited from a hospital in Seoul and they had cochlear implants for longer than two years(M = 4.7). NH children were randomly sampled. All participants were asked to respond to a questionnaire and rate their music preferences. The results showed that while 58.33% of CI recipients preferred child voices, 71.61% of NH children preferred adult voices. For preferred frequency range, both of CI and NH groups preferred higher frequency. With regard to preferred musical styles, while CI children gave the highest rating(M = 3.58) to Korean folk songs on the 5-point Likert scale, popular music and TV and soundtrack music were found to be the highest-rated styles(M = 3.83) for NH children. These results indicate that self-rated music preferences of CI recipients may be affected by a musical environment that these children have experienced. The results also implies that the provision of appropriate musical input and experiences would significantly affect future music preferences and appreciation of CI children. Providing useful information about music preferences of CI recipients in comparison with NH children, this study has significant implications for further studies on this population.