.png){kind=icon}
Return
A machine-learning algorithm for wind gust prediction
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cageo.2011.03.004.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Physical damage to property and crops caused by unanticipated wind gusts is a well understood phenomenon. Predicting its occurrence continues to be a challenge for meteorologists and climatologists. Various approaches to gust occurrence model building have been proposed. The very nature of the event is problematic because of its brief duration following a rapid change of state in wind velocity that immediately precedes it. Events classified as wind gusts have a typical duration of less than 20 s and are often much shorter. The rapidly accelerating wind velocity preceding them is often not apparent until the gust occurs. They come quickly, occur suddenly, and then end as abruptly as they began. Observations of 2000 gust events were made during the research to which this paper refers. These observations indicated a mean interval of 3.2 min between the beginning and end of wind velocity change and a noticeable linear progression in the acceleration pattern. It was also noted that state changes regularly occur, often over only seconds in time. In combination, these factors pose both a sampling and a data interpretation challenge, making reliable prediction difficult. This paper describes some new research undertaken to investigate methods of wind gust measurement and prediction. In particular, a machine-learning approach is taken to determine a satisfactory analytical process and to produce meaningful and useful results. An algorithm for use with real-time climate data collection and analysis is proposed, with a description of its implementation. Real-time data sampling provides input for this study using terrestrial sensor telemetry. Near-ground truth data are recorded independent of geostrophic upper atmosphere conditions. (C) 2011 Elsevier Ltd. All rights reserved.
Keywords:
Wind velocity modeling
Wind gust prediction
Machine-learning algorithms
Geostatistics
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
C
IF:
4.4
Papers:
5.0K
Citations:
1.5W
Organization
Cited Papers
Arsenic removal from aqueous solutions by adsorption using novel MIL-53(Fe) as a highly efficient adsorbent.png)
RSC Advances
RSC AdvancesIF0
PREAI
ShareSave
ShareSaveA comparison of prediction accuracy, complexity, and training time of thirty-three old and new classification algorithmsMACHINE LEARNING
MACHINE LEARNINGIF2.9
PREAI
ShareSaveShareSave
ShareSave
View more