Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

DSD-INT 2019 A model-oriented interlink among platforms of IoT, AI, and Delft-FEWS - Chang

19 views

Published on

Presentation by Che Hao Chang , Taipei Tech - National Taipei University of Technology, at the Delft-FEWS User Days, during Delft Software Days - Edition 2019. Thursday, 7 November 2019, Delft.

Published in: Software
  • Be the first to comment

  • Be the first to like this

DSD-INT 2019 A model-oriented interlink among platforms of IoT, AI, and Delft-FEWS - Chang

  1. 1. A model oriented interlink among platforms of IoT, AI, and Delft-FEWS HPC Big Data Cloud Network Che-Hao Chang, Associate Professor, National Taipei University of Technology chchang@ntut.edu.tw Chih-Tsung Hsu, Shiang-Jen Wu, Researcher, NCHC Jhih-Cyuan Shen, CEO, FondUs Inc. DSD-INT 2019 in Delft, 2019/11/07
  2. 2. Uncertainties in Hydrosytem
  3. 3. 3 Max 7mm/hr Rainfall center not presented by gauges Max 22.5mm/hr in 10min 19.21mm/hr in an hour Drainage block Radar Observation Gauge Tainan 2018 0906 Tainan City High Spatial Uncertainty
  4. 4. 4 Limited Gauges Distribution 4 1.3 km resolution of radar rainfall 2017 2016 2015
  5. 5. Unit :mm/10min 17km High Intensity in Short Duration C1V410 C1V430 C1V440 C1V420
  6. 6. EXTREME RAINFALL IN KAOHSIUNG TYPHOON FANAPI 20100919 5 hour : 539mm Max rainfall:124mm/hr 5 hour : 289.5mm Max rainfall:87mm/hr 5 hour : 280mm Max rainfall:78mm/hr 5 hour : 477mm Max rainfall :124.5mm/hr C1V410 C1V430 C1V440 C1V420 岡山 左營 鳳山 鳳雄
  7. 7. NCDR watch Drought And Water Distribution Hsinchu Science Park: 2016 Revenu over 30,000 million USD Boston Red Sox 2016 payroll : 208 million USD
  8. 8. Sensors & Model Simulation Deal with Uncertainty: Sensors, sensors, and sensors From points to area
  9. 9. Flood Grid at An-Jhong 0 2 0 40 m
  10. 10. 0 0,1 0,2 0,3 0,4 0,5 Depth(m) Simulated and observed flood height at An-Jhong Observed Simulated
  11. 11. Flood Grid at Chao-Huang-Gong Station 0 20 40 m
  12. 12. 0 0,1 0,2 0,3 0,4 0,5 Depth(m) Simulated and observed flood height at Chao-Huang-Gong Observed Simulated
  13. 13. IoT Techs bring us.. ▪ Energy Harvesting technology ▪ Low power chip ▪ Low power WAN ▪ New Battery technology ▪ Industrial standard chip but consumer product price.
  14. 14. • https://www.theweathernetwor k.com/us/news/articles/app- and-volunteer-army-improving- tidal-forecasts/90256
  15. 15. 三崁店 Microwave + LoRa Relay 北安 Microwave + LoRa Relay 鯤鯓 Microwave + LoRa Relay 鹿耳門 Microwave NB-IoT/3G/4G Cell Tower Sensors (Within LoRa Range) 安污Control Center Sensors (Outside LoRa Range) Only Channel
  16. 16.  Every 1 hr provide next 3 hrs flood nowcast  Every 2 hr provide next 6-8 hrs flood nowcast . OPERATIONAL FLOOD EARLY WARNING SYSTEM FEWS_TAIWAN
  17. 17. Interlink among Platforms Model update: Smart Model Management Quick response: AI assist
  18. 18. • > 1000 rain gauges • > 400 water level gauges • >1000 flood sensors and even more – (x,y,depth,t) from voice, video, messages • Smart water meters
  19. 19. 28/35 Radar Rainfall Storm Generator
  20. 20. 29/35 Event Select Extract Radar Rainfall By data and location Radar Rainfall historical DB 2006~2018 Radar Rainfall Storm Generator
  21. 21. 30/35 Historical Storm Events Event Select Radar Rainfall Storm Generator
  22. 22. 31/35 r Radar Rainfall Event Select Historical Events Spatiotemporal Randomess 0 200 400 600 800 1000 1200 1400 1600 1800 2000 1 2 3 4 5 6 7 8 9 10 Rainfall(mm) No of study gird EV2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulativedimensionlessrainfall Dimensionless time PT1 PT2 PT3 PT4 PT5 PT6 PT7 PT8 PT9 PT10 Areal average Bias Simulatedareal average of cumulative dimensionlessrainfall Simulated bias of cumulative dimensionless rainfall Simulated rainfall depthat 10 studygrids Simulated storm patternsat 10 studygrids Simulated hyetographsat 10 studygrids 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulativedimensionlessrainfall Dimensionless time -0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0 0.2 0.4 0.6 0.8 1 Biasofdimensionlessrainfall Dimensionless time PT1 PT2 PT3 PT4 PT5 PT6 PT7 PT8 PT9 PT10 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumulativedimensionless rainfall Dimensionlesstime RT1 RT2 RT3 RT4 RT5 RT6 RT7 RT8 RT9 RT10 0 500 1000 1500 2000 2500 3000 1 2 3 4 5 6 7 8 9 10 Rainfall(mm) No of study grid 0 10 20 30 40 50 60 70 80 90 100 1 7 13 19 25 31 37 43 49 55 61 67 7379 85 91 Rainfall(mm) Time (hour) 數列2 數列3 數列4 數列5 數列6 數列7 數列8 數列9 數列10 數列11 … … … … Multi-Scenario Generation Storm Generator Multi_Scenario storms DB
  23. 23. 32/35 TWCC Database of multi- spatiotemporal storms Arrange the VM for simulations … … … … Database of inundation simulations 2D rainfall-runoff-inundation simulation … … … …
  24. 24. Framework
  25. 25. 34/35 Storm Generator

×