本文選題：房屋震害損失評估 + 云計算��； 參考：《中國地震局工程力學研究所》2017年碩士論文

【摘要】：房屋震害損失評估是地震災害損失評估非常重要的一環(huán),對震后災區(qū)救援和復建具有很大的指導意義和參考價值。傳統(tǒng)的房屋震害損失評估采取抽樣方法來采集數(shù)據(jù),然后基于這些具有一定代表性的數(shù)據(jù)來進行評估,然而,用樣本數(shù)據(jù)來代替總體數(shù)據(jù)是一種“無奈之舉”,會在一定程度上削弱評估結(jié)果的準確性。隨著信息時代的到來,各個行業(yè)領域的信息量快速增長,數(shù)據(jù)類型越來越復雜,大數(shù)據(jù)的概念就是在這樣的背景下被提出的。大數(shù)據(jù)不僅僅是數(shù)據(jù)量巨大,同時還兼具異構(gòu)性和價值性。但是要想發(fā)掘大數(shù)據(jù)所蘊藏的價值性,需要便捷快速、經(jīng)濟的運算工具。云計算是一種通過共享云端資源池來實現(xiàn)低成本高運算的計算模式,目前,公認是處理大數(shù)據(jù)的最佳利器。地震災害評估方面的數(shù)據(jù)也在急劇增長中,面對這種增長,傳統(tǒng)的計算和數(shù)據(jù)存儲方法在處理速度上越發(fā)顯得捉襟見肘�！吨泄仓醒雵鴦赵宏P于進一步加強城市規(guī)劃建設管理工作的若干意見》指出要“推進城市智慧管理。加強城市管理和服務體系智能化建設,促進大數(shù)據(jù)、物聯(lián)網(wǎng)、云計算等現(xiàn)代信息技術(shù)與城市管理服務融合,提升城市治理和服務水平�！钡卣馂暮υu估也需要推進智慧管理,將現(xiàn)代信息技術(shù)手段引入到地震災害評估等相關領域,以此得到更好的發(fā)展。論文主要做了以下幾項工作:1)綜述大數(shù)據(jù)、云計算在地震數(shù)據(jù)處理領域運用的國內(nèi)外發(fā)展現(xiàn)狀。通過大量的資料查詢以及分析對比工作,最終選定本文的云開發(fā)平臺為Hadoop。同時結(jié)合本文數(shù)據(jù)存儲以及計算速度的需求,明確了本文采用的云計算三項技術(shù):MapReduce編程模型、HDFS分布式文件系統(tǒng)以及HBase非關系型數(shù)據(jù)庫,并對這三項相關技術(shù)進行了必要的介紹。2)詳細分析了目前傳統(tǒng)房屋震害損評估方法存在的不足,結(jié)合房屋震害數(shù)據(jù)采集手段不斷進步的發(fā)展趨勢,設計了一種理想化的基于大數(shù)據(jù)應用云計算技術(shù)評估詳實房屋震害損失的方法,重點是用全數(shù)據(jù)參與評估。并且完成以下幾部分設計:a)數(shù)據(jù)存儲方面,設計出合理的便于存儲以及進行數(shù)據(jù)操作的HBase數(shù)據(jù)表結(jié)構(gòu)。并且完成了數(shù)據(jù)批量錄入HBase的程序設計。b)數(shù)據(jù)處理部分:將整個評估過程進行合理的拆分,設計出高效可靠的算法流程,并通過MapReduce程序來實現(xiàn)。3)將大量數(shù)據(jù)導入HBase數(shù)據(jù)庫,在數(shù)據(jù)庫相同的條件下,分別運用云計算集群和傳統(tǒng)單機模式計算房屋震害損失值,比較本評估方法運用兩種計算工具的計算速度。
[Abstract]:Building damage assessment is a very important part of earthquake disaster loss assessment, which has great guiding significance and reference value for disaster relief and reconstruction after earthquake.The traditional method of building damage assessment is to collect data by sampling method, and then evaluate it based on these representative data. However, it is "helpless" to replace the overall data with sample data.To some extent, the accuracy of the evaluation results will be weakened.With the arrival of the information age, the amount of information in various industries is increasing rapidly, and the data types are becoming more and more complex. Big data's concept is put forward under this background.Big data is not only a huge amount of data, but also heterogeneity and value.But in order to explore the value of big data, need convenient and rapid, economic computing tools.Cloud computing is a kind of computing mode which can achieve low cost and high operation by sharing cloud resource pool. At present, cloud computing is recognized as the best weapon to deal with big data.Data on earthquake disaster assessment are also growing dramatically, and in the face of this growth,The traditional methods of calculation and data storage are more and more overstretched in processing speed. "some opinions of the CPC Central Committee and the State Council on further strengthening the management of urban planning and construction" pointed out that "promoting urban intelligent management".Strengthen the intelligent construction of urban management and service system, promote the integration of modern information technology such as big data, Internet of things, cloud computing and urban management services, and improve the level of urban governance and services. "Earthquake disaster assessment also needs to promote intelligent management and introduce modern information technology into earthquake disaster assessment and other related fields in order to get a better development.This paper summarizes the development of big data, cloud computing in the field of seismic data processing at home and abroad.Finally, the cloud development platform of this paper is chosen as Hadoop through a lot of data query and analysis and comparison work.At the same time, according to the requirement of data storage and computing speed in this paper, three technologies of cloud computing, namely: MapReduce programming model, HDFS distributed file system and HBase non-relational database, are defined in this paper.At the same time, the necessary introduction of these three related technologies. 2) analyzes in detail the shortcomings of the traditional methods for evaluating the damage of buildings, and combines with the developing trend of the means of collecting data on the earthquake damage of buildings.An idealized method based on big data's application of cloud computing technology is designed to evaluate the damage loss of detailed buildings, with the emphasis on full data participation.At the same time, the following several parts are completed to design the HBase data table structure which is convenient to store and operate.And completed the program design of data batch input into HBase. B) data processing part: the whole evaluation process is divided reasonably, the efficient and reliable algorithm flow is designed, and a large amount of data is imported into HBase database through MapReduce program.Under the same database condition, the cloud computing cluster and the traditional single-machine model are used to calculate the damage value of buildings, and the calculation speed of the two computing tools is compared in this evaluation method.
【學位授予單位】：中國地震局工程力學研究所
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：P315.9;TP311.13

【參考文獻】

相關期刊論文 前10條

1 龔強;李萌;;基于大數(shù)據(jù)用云計算方法對地震房屋損失評估的研究[J];信息技術(shù);2017年05期

2 龔強;李萌;;一種基于大數(shù)據(jù)云計算的地震房屋損失評估模型[J];防災減災學報;2017年01期

3 陳會忠;;地震大數(shù)據(jù)思維[J];城市與減災;2016年02期

4 龔強;;云計算關鍵技術(shù)之編程模型認知研究[J];信息技術(shù);2015年01期

5 李青云;余文;;關系型數(shù)據(jù)庫到H Base的轉(zhuǎn)換設計[J];信息網(wǎng)絡安全;2015年01期

6 邢揚;張國江;;橋梁震害分析與抗震設計新方法研究[J];山西建筑;2015年01期

7 屈佳;鄭蕊;王寧;;地震行業(yè)“大數(shù)據(jù)”應用探討[J];城市與減災;2014年04期

8 高源;;NoSQL非關系型數(shù)據(jù)庫的發(fā)展和應用研究[J];計算機光盤軟件與應用;2014年05期

9 崔忠偉;左羽;韋萍萍;熊偉程;;基于云計算的數(shù)字圖書館服務平臺架構(gòu)設計[J];物聯(lián)網(wǎng)技術(shù);2014年02期

10 何清;;大數(shù)據(jù)與云計算[J];科技促進發(fā)展;2014年01期

相關碩士學位論文 前5條

1 劉飛;基于云計算的分布式存儲系統(tǒng)的研究和應用[D];西安工業(yè)大學;2012年

2 李崇欣;分布式數(shù)據(jù)庫HBase快照的設計與實現(xiàn)[D];浙江大學;2011年

3 翟永東;Hadoop分布式文件系統(tǒng)（HDFS）可靠性的研究與優(yōu)化[D];華中科技大學;2011年

4 李利;城市建筑物震害及震害經(jīng)濟損失預測方法研究[D];大連理工大學;2009年

5 陳洪富;城市房屋建筑裝修震害損失評估方法研究[D];中國地震局工程力學研究所;2008年

，

本文編號：1735925