發(fā)布時間：2018-06-17 00:07

  本文選題：農(nóng)業(yè)大數(shù)據(jù)平臺 + 數(shù)據(jù)可視化　； 參考：《東北農(nóng)業(yè)大學(xué)》2017年碩士論文

【摘要】：隨著農(nóng)業(yè)現(xiàn)代信息化進(jìn)程的不斷發(fā)展以及農(nóng)業(yè)種植、畜牧業(yè)、漁業(yè)、農(nóng)產(chǎn)品加工、氣象等數(shù)據(jù)的不斷積累,農(nóng)業(yè)數(shù)據(jù)正以前所未有的速度不斷增長并形成了海量數(shù)據(jù)。我國農(nóng)業(yè)領(lǐng)域數(shù)據(jù)具有數(shù)據(jù)實時性強(qiáng)、維度高、數(shù)據(jù)存儲分散、難于綜合分析等特性,一方面是因為我國農(nóng)業(yè)結(jié)構(gòu)復(fù)雜,農(nóng)業(yè)數(shù)據(jù)涉及多個領(lǐng)域,另一方面農(nóng)業(yè)數(shù)據(jù)又容易受到地理環(huán)境、土壤、天氣、作物、病蟲害等的影響。這些數(shù)據(jù)的有效利用需要相應(yīng)的大數(shù)據(jù)平臺作為支撐,大數(shù)據(jù)平臺可以整合農(nóng)業(yè)領(lǐng)域的數(shù)據(jù),提供查詢、下載、上傳、可視化等功能;平臺的數(shù)據(jù)挖掘方法可以挖掘隱藏在農(nóng)業(yè)數(shù)據(jù)中的知識,發(fā)現(xiàn)規(guī)律;大數(shù)據(jù)平臺還可以為農(nóng)業(yè)工作者提出決策意見和指導(dǎo)建議。所以開發(fā)具有以上功能的農(nóng)業(yè)大數(shù)據(jù)平臺具有重要現(xiàn)實意義。本文根據(jù)農(nóng)業(yè)大數(shù)據(jù)的性質(zhì),分析了農(nóng)業(yè)大數(shù)據(jù)平臺的主要技術(shù),對農(nóng)業(yè)數(shù)據(jù)分析和可視化工具進(jìn)行了較深入的探討,提出了基于改進(jìn)的譜聚類算法,并搭建了具有挖掘功能的農(nóng)業(yè)大數(shù)據(jù)平臺。論文主要工作如下:在農(nóng)業(yè)數(shù)據(jù)收集方面我們通過中國統(tǒng)計年鑒、高校的實驗數(shù)據(jù)和相關(guān)農(nóng)業(yè)網(wǎng)站下載了一定量的農(nóng)業(yè)數(shù)據(jù),同時還使用網(wǎng)絡(luò)爬蟲技術(shù)抓取了一些農(nóng)業(yè)相關(guān)網(wǎng)站數(shù)據(jù),并利用這些數(shù)據(jù)建立了農(nóng)業(yè)數(shù)據(jù)庫。在農(nóng)業(yè)數(shù)據(jù)平臺數(shù)據(jù)挖掘和可視化兩項關(guān)鍵技術(shù)研究方面,我們提出了針對農(nóng)業(yè)大數(shù)據(jù)的基于閔可夫斯基測量相似程度的改進(jìn)譜聚類算法,在UCI數(shù)據(jù)集中的seeds和soybean數(shù)據(jù)集的仿真實驗結(jié)果表明論文提出的算法在聚類精度和運(yùn)算速度上都有一定的提高。我們還使用多維數(shù)據(jù)的數(shù)據(jù)可視化技術(shù),實現(xiàn)平臺的數(shù)據(jù)可視化功能。在平臺開發(fā)方面我們根據(jù)平臺需求給出了系統(tǒng)設(shè)計方案并使用J2EE相關(guān)技術(shù)實現(xiàn)了平臺的全部功能,在系統(tǒng)測試方面我們利用農(nóng)業(yè)機(jī)械、畜牧業(yè)數(shù)據(jù)進(jìn)行了實驗仿真。本文開發(fā)的農(nóng)業(yè)大數(shù)據(jù)平臺,用戶界面友好使用簡單,在數(shù)據(jù)收集方面我們通過網(wǎng)絡(luò)爬蟲模塊,實現(xiàn)了數(shù)據(jù)自動獲取和存儲。除可以實現(xiàn)目前農(nóng)業(yè)數(shù)據(jù)平臺查詢、上傳、下載的功能外,還具有性能優(yōu)越的數(shù)據(jù)挖掘模塊和簡單易懂的數(shù)據(jù)可視化模塊。平臺設(shè)計合理且實用。農(nóng)業(yè)大數(shù)據(jù)關(guān)鍵技術(shù)的研究和平臺的開發(fā)對農(nóng)業(yè)信息化和智慧農(nóng)業(yè)的發(fā)展具有重要的參考價值和推動作用。
[Abstract]:With the continuous development of modern agricultural information process and the continuous accumulation of agricultural planting, animal husbandry, fishery, processing of agricultural products and meteorological data, agricultural data is growing at an unprecedented rate and forming massive data. The data of agricultural field in our country have high real-time data, high dimension, scattered data storage and difficult to integrate. On the one hand, it is because of the complex agricultural structure in China, the agricultural data are involved in many fields, and on the other hand, the agricultural data are easily affected by the geographical environment, soil, weather, crops, diseases and pests. The effective utilization of these data needs the corresponding large data platform as support, and the large data platform can integrate the agricultural field. Data, providing the functions of query, downloading, uploading, visualization and so on. The data mining method of the platform can discover the knowledge hidden in the agricultural data and discover the rules. The large data platform can also provide advice and advice for the agricultural workers. Therefore, it is of great practical significance to develop a large agricultural data platform with the above functions. According to the nature of large agricultural data, the main technology of agricultural data platform is analyzed, and the agricultural data analysis and visualization tools are discussed. The improved spectral clustering algorithm is proposed and a large agricultural data platform with mining function is set up. The main work of this paper is as follows: in the field of agricultural data collection, I We have downloaded a certain amount of agricultural data through the Chinese Statistical Yearbook, the experimental data of the University and the related agricultural websites. At the same time, we also use the web crawler technology to capture some agricultural related website data, and use these data to establish the agricultural database. In the field of data mining and visualization of the agricultural data platform, two key technologies are studied. We propose an improved spectral clustering algorithm based on the similarity degree of Minkowski measurement for large agricultural data. The simulation results of seeds and soybean data sets in the UCI data set show that the proposed algorithm has a certain improvement in clustering accuracy and computing speed. We also use multidimensional data visualization techniques. In the aspect of platform development, we give the system design scheme according to the platform requirements and use J2EE related technology to realize all the functions of the platform. In the system testing, we use agricultural machinery and animal husbandry data to imitate true. The agricultural big data platform and the user community developed in this paper. The face is friendly and simple. In data collection, we have realized data acquisition and storage through the network crawler module. In addition to the functions of agricultural data platform query, uploading and downloading, it also has excellent data mining module and simple and easy to understand data visualization module. The platform is designed to be reasonable and practical. The key technology research and platform development of big data has important reference value and impetus to the development of agricultural informatization and intelligent agriculture.
【學(xué)位授予單位】：東北農(nóng)業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：F323.3

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 郭風(fēng)成;李參海;李宗春;王華斌;;引入歐氏距離的各向異性擴(kuò)散相干斑抑制[J];中國圖象圖形學(xué)報;2017年03期

2 王東杰;李哲敏;張建華;許世衛(wèi);;農(nóng)業(yè)大數(shù)據(jù)共享現(xiàn)狀分析與對策研究[J];中國農(nóng)業(yè)科技導(dǎo)報;2016年03期

3 藺旭東;周軍鋒;劉佳;;資源關(guān)聯(lián)性大數(shù)據(jù)分析在農(nóng)業(yè)生態(tài)環(huán)境保護(hù)中的應(yīng)用[J];中國農(nóng)業(yè)資源與區(qū)劃;2016年02期

4 朱倩;王華麗;;基于中間件的新疆農(nóng)業(yè)科學(xué)數(shù)據(jù)共享平臺的體系構(gòu)建[J];天津農(nóng)業(yè)科學(xué);2016年01期

5 李學(xué)龍;龔海剛;;大數(shù)據(jù)系統(tǒng)綜述[J];中國科學(xué):信息科學(xué);2015年01期

6 張浩然;李中良;鄒騰飛;魏旭陽;楊國才;;農(nóng)業(yè)大數(shù)據(jù)綜述[J];計算機(jī)科學(xué);2014年S2期

7 阮懷軍;封文杰;唐研;趙佳;李道亮;;農(nóng)業(yè)信息化建設(shè)的實證研究——以山東省為例[J];中國農(nóng)業(yè)科學(xué);2014年20期

8 李秀峰;陳守合;郭雷風(fēng);;大數(shù)據(jù)時代農(nóng)業(yè)信息服務(wù)的技術(shù)創(chuàng)新[J];中國農(nóng)業(yè)科技導(dǎo)報;2014年04期

9 肖進(jìn)勝;饒?zhí)煊?賈茜;宋金鐘;易本順;;基于圖切割的拉普拉斯金字塔圖像融合算法[J];光電子.激光;2014年07期

10 宋振維;;探究分析JAVA語言的開發(fā)平臺及J2EE編程技術(shù)[J];計算機(jī)光盤軟件與應(yīng)用;2014年06期

相關(guān)博士學(xué)位論文 前1條

1 郭雷風(fēng);面向農(nóng)業(yè)領(lǐng)域的大數(shù)據(jù)關(guān)鍵技術(shù)研究[D];中國農(nóng)業(yè)科學(xué)院;2016年

相關(guān)碩士學(xué)位論文 前7條

1 馮陽;大數(shù)據(jù)技術(shù)在農(nóng)技推廣中的應(yīng)用研究[D];中國農(nóng)業(yè)科學(xué)院;2016年

2 張健;基于WebGIS的農(nóng)業(yè)地理數(shù)據(jù)可視化技術(shù)研究及應(yīng)用[D];浙江大學(xué);2015年

3 洪禮;新型農(nóng)業(yè)平臺的研究與開發(fā)[D];浙江理工大學(xué);2015年

4 張亞平;譜聚類算法及其應(yīng)用研究[D];中北大學(xué);2014年

5 張宏東;EM算法及其應(yīng)用[D];山東大學(xué);2014年

6 鄭晨俊;可重構(gòu)嵌入式CPU平臺JAVA開發(fā)及其在農(nóng)業(yè)物聯(lián)網(wǎng)中的應(yīng)用[D];復(fù)旦大學(xué);2014年

7 藍(lán)龍;半監(jiān)督非負(fù)矩陣分解算法及其在文本聚類中的應(yīng)用[D];國防科學(xué)技術(shù)大學(xué);2012年

，

本文編號：2028651