The Research on Summarizing Flood Forecasting Messages

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

NSC102-2221-E019-063

Translated Name/計畫中文名

洪氾預警資訊之摘要化研究

Project Coordinator/計畫主持人

Ya-Hui Chang

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Department of Computer Science and Engineering

Website

https://www.grb.gov.tw/search/planDetail?id=3108968

Year

2013

Start date/計畫起

01-08-2013

Expected Completion/計畫迄

31-07-2014

Bugetid/研究經費

554千元

ResearchField/研究領域

資訊科學--軟體
防災工程

洪氾預警系統可在颱風或大雨來臨時，利用當下可得的氣象資料，計算出未來 1-3 小時的地區水深，並在達到警戒範圍時即時提出警告疏散民眾。但是，目前預警資訊大多是透過地圖標示並讓使用者透過瀏覽器觀看，在淹水嚴重的情況下，大量的預警資訊無法即時且清楚地顯示。本計畫希望將預警資訊進行摘要化的處理。我們提出一種創新的概念，稱做「基於地標的警示訊息」。所謂的「地標」，是指大家熟知的機關學校或建築物，所摘要的警示訊息如下：「昌隆國小北方約 5 公里處會發生大約 0.55 平方公里的淹水」。此類摘要訊息可快速地傳送，並提供更有意義的資訊。本計畫希望進行三年的研究，其分年的目標如下：(1)第一年度，我們假設每個淹水點的面積大致相同，所以只需處理點的座標和個數。我們基於關聯式資料庫的技術，定義相關的資料表格和撰寫對應的 SQL 查詢句來推算摘要資訊。(2)第二年度，我們考慮一般的狀況，也就是淹水點的面積具有差異性。我們基於 Voronoi cell 的觀念，以地標為基準切割空間，然後建立 R-tree 索引，進行空間連結的運算，來計算每個地標附近的淹水點。我們也會研究如何利用資料探勘中分群的技術，協助將淹水點分群。(3)在第三年度，我們將進行廣泛的實驗，來評估不同做法的優劣。我們也將實做手機 App，以便使用者隨時隨地皆可迅速取得預警資訊。若本計畫可以順利執行，不僅在防災和地理資訊處理的學術領域有所貢獻，對政府防災系統的建置也會有實質的助益。Building flood forecasting systems is an important research issue. Such system can predict the water level for the next one to three hours, and provide early warning messages. However, currently the warning message is displayed on the map. It is not efficient and not clear when the flooded situation is severe. In this project, we wish to summarize the warning messages based on the novel idea of “landmarks”. Landmarks refer to those buildings or places such as schools or hospitals. Landmark-based warning messages can be efficiently broadcast, and can provide meaningful information. We plan to carry out this project for three years to investigate different kind of techniques as follows: (1) In the first year, we will mainly base on the techniques of relational databases to represent the positions and numbers of flooded locations, and use SQL to retrieve summarized information. (2) In the second year, we will utilize the idea of Voronoi cells and the R-tree index to identify the flooded locations near each landmark. (3) In the last year, we will perform extensive experiments to evaluate each approach, and implement the Web-based systems and Apps for mobile phones. If this plan can be successfully carried out, it will contribute a lot to the research for spatial information processing, and be also beneficial to the implementation of flood forecasting systems.

Keyword(s)

洪氾預警系統
地標
關聯式資料庫
索引
flood forecasting system
landmark
relational database
index

DSpace CRIS

The Research on Summarizing Flood Forecasting Messages

Details

Description