• 
    

    
    

      99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

      基于近鄰傳播算法的茶園土壤墑情傳感器布局優(yōu)化

      2019-05-11 07:03:00張嫚嫚江朝暉蔣躍林
      關(guān)鍵詞:墑情布局含水率

      張 武,張嫚嫚,洪 汛,江朝暉,蔣躍林

      ?

      基于近鄰傳播算法的茶園土壤墑情傳感器布局優(yōu)化

      張 武1,張嫚嫚1,洪 汛1,江朝暉1※,蔣躍林2

      (1.安徽農(nóng)業(yè)大學(xué)信息與計(jì)算機(jī)學(xué)院,合肥 230036;2.安徽農(nóng)業(yè)大學(xué)資源與環(huán)境學(xué)院,合肥 230036)

      針對節(jié)水灌溉的土壤墑情傳感器布局問題,提出了基于近鄰傳播算法(affinity propagation,AP聚類算法)的優(yōu)化布局策略。策略在保證茶園傳感網(wǎng)絡(luò)全覆蓋的基礎(chǔ)上,實(shí)時采集試驗(yàn)區(qū)各節(jié)點(diǎn)的土壤墑情數(shù)據(jù),構(gòu)建節(jié)點(diǎn)土壤含水率的相似度矩陣,并迭代計(jì)算各節(jié)點(diǎn)的吸引度和歸屬度值,得出聚類數(shù)和聚類中心。結(jié)果表明,采用AP聚類算法對試驗(yàn)區(qū)域傳感器進(jìn)行優(yōu)化布局,優(yōu)化了傳感器數(shù)量和位置,使傳感器數(shù)量由25個減少到2個。在試驗(yàn)區(qū)隨機(jī)采集土壤相對含水率,經(jīng)驗(yàn)證,聚類中心節(jié)點(diǎn)的土壤相對含水率與試驗(yàn)區(qū)平均值相近,相對偏差近為0.76%,表明聚類中心節(jié)點(diǎn)的土壤墑情數(shù)據(jù)具有代表性。該方法有效降低了數(shù)據(jù)的冗余度,節(jié)約了系統(tǒng)成本。

      墑情;傳感器;聚類算法:優(yōu)化布局;AP

      0 引 言

      農(nóng)業(yè)節(jié)水灌溉系統(tǒng)通過分布于農(nóng)田的傳感器采集土壤墑情信息,合理的傳感器選擇和優(yōu)化布局對土壤墑情的準(zhǔn)確獲取起著重要作用[1]。傳感器布局的優(yōu)化問題在航空航天領(lǐng)域、結(jié)構(gòu)損傷探測和機(jī)械故障診斷分析等方面獲得了廣泛應(yīng)用[2-5]。

      目前人們對于傳感器布局研究的重點(diǎn)集中于布點(diǎn)模型和覆蓋算法,即通過不同的算法模型,從基站數(shù)量、網(wǎng)絡(luò)連通性以及在有障礙物環(huán)境中的最小覆蓋度等方面開展優(yōu)化研究。劉麗萍[6]在布點(diǎn)模型方面研究了隨機(jī)規(guī)則布點(diǎn)模型和泊松分布布點(diǎn)模型,利用臨界傳感器密度來確保完全覆蓋。在覆蓋算法方面,主要涉及遺傳算法、貪婪策略方法和粒子群算法等。Lin等[7]從傳感器網(wǎng)絡(luò)能量最優(yōu)的角度,運(yùn)用模擬退火算法對傳感器布局進(jìn)行了優(yōu)化設(shè)計(jì);Wang等[8]針對有障礙物的傳感區(qū)域,以實(shí)現(xiàn)最小覆蓋為目標(biāo)優(yōu)化了傳感網(wǎng)絡(luò);趙偉霞等[9]研究了土壤含水率的時間穩(wěn)定性問題,分析了將土壤水分傳感器直接布設(shè)在代表平均土壤含水率的點(diǎn)時可選擇的測點(diǎn)數(shù)量與灌水均勻系數(shù)和土層深度的關(guān)系;蔣杰等[10]設(shè)計(jì)了一種基于目標(biāo)區(qū)域Voronoi劃分的集中式近似算法和基于最小生成樹的連通算法;李飚等[11]針對土壤墑情傳感器布點(diǎn)問題提出了一種基于Delaunay三角剖分的傳感器布點(diǎn)方法。

      有關(guān)農(nóng)田土壤墑情傳感器布局的研究主要側(cè)重于網(wǎng)絡(luò)覆蓋的優(yōu)化算法方面,對于土壤墑情數(shù)據(jù)的冗余度問題尚缺少相關(guān)的研究。通常,農(nóng)田土壤的含水率受降雨、地形、土壤特性、植被分布和微立地條件等多種因素影響[12-15],同質(zhì)地土壤在平面和深度上實(shí)際并非完全均質(zhì),大多數(shù)土壤特性均是空間位置的函數(shù),在同質(zhì)地不同位置處存在著差異[16]。在農(nóng)田土壤墑情傳感器的布局優(yōu)化過程中不僅要實(shí)現(xiàn)傳感網(wǎng)絡(luò)的全覆蓋,還應(yīng)考慮土壤墑情數(shù)據(jù)的相似度和差異性問題,以減少數(shù)據(jù)的冗余度為目標(biāo)合理優(yōu)化傳感器布局。

      本文以茶園土壤墑情為研究對象,研究土壤墑情信息的差異性,在網(wǎng)絡(luò)全覆蓋的基礎(chǔ)上基于近鄰傳播算法(affinity propagation,AP聚類算法)通過計(jì)算土壤節(jié)點(diǎn)含水率的相似度矩陣實(shí)現(xiàn)墑情數(shù)據(jù)的聚類,優(yōu)化傳感器布局,降低數(shù)據(jù)的冗余度和系統(tǒng)成本。

      1 傳感器布局

      1.1 傳統(tǒng)布局方法

      土壤墑情傳感器一般多采用均勻分布的方式,即將傳感器分別放在農(nóng)田網(wǎng)格(一般為正四邊形)的中心點(diǎn),以保證傳感網(wǎng)絡(luò)的全覆蓋。但這種布局方式存在一定的缺陷,如果傳感器均勻分布的太稀疏,則無法實(shí)現(xiàn)傳感器網(wǎng)絡(luò)的全覆蓋;相反,如果傳感器分布的過于密集,則會使傳感器覆蓋范圍中傳感區(qū)域重疊過多,導(dǎo)致成本的上升和資源的浪費(fèi)。

      在傳感器布局中,不僅需要滿足傳感器的多種性能約束,如系統(tǒng)能量損耗、覆蓋精度、信號的完備性等,還需要考慮數(shù)據(jù)的冗余度,因此,傳感器的布局需要結(jié)合數(shù)據(jù)的相似度和差異性進(jìn)行優(yōu)化布局。

      1.2 基于AP布局

      本文在茶園土壤墑情傳感器的布局過程中,首先在目標(biāo)區(qū)域按照四邊形進(jìn)行布局,即將傳感器布置于各四邊形節(jié)點(diǎn)上,滿足全覆蓋的目標(biāo);其次,基于AP聚類算法構(gòu)建各節(jié)點(diǎn)土壤墑情的相似度矩陣,計(jì)算各節(jié)點(diǎn)的吸引度和歸屬度值,通過迭代計(jì)算得出目標(biāo)區(qū)域的土壤墑情的聚類中心,以聚類中心為基準(zhǔn),剔除冗余傳感節(jié)點(diǎn),優(yōu)化傳感器布局[17-18]。

      1.2.1 AP聚類算法簡介

      AP聚類算法是一種基于數(shù)據(jù)點(diǎn)間的“消息傳遞”的聚類算法[19]。該算法不需要先確定聚類的數(shù)目,而是把所有的數(shù)據(jù)點(diǎn)都看成潛在意義上的聚類中心,然后通過節(jié)點(diǎn)之間的消息傳遞找到最合適的聚類中心。該算法是一種確定性的聚類算法,多次獨(dú)立運(yùn)行的聚類結(jié)果穩(wěn)定,已經(jīng)被應(yīng)用到多個領(lǐng)域,如圖像識別、圖像分割,圖像檢索及文本挖掘等,取得了較好的效果[20-22]。

      1.2.2 AP聚類算法基本原理及流程

      設(shè)數(shù)據(jù)樣本集{1,2, …, x},AP聚類算法用負(fù)的歐氏距離表示數(shù)據(jù)點(diǎn)和(共個)之間的相似度(,)。

      個數(shù)據(jù)點(diǎn)之間構(gòu)成×的相似度矩陣=[(,)]×n,聚類中心由對角線上元素的數(shù)值(即偏向參數(shù),)決定。值影響聚類結(jié)果,越大聚類數(shù)目越多,值越小聚類數(shù)目越少,一般取對角線上值的中值[23]。

      數(shù)據(jù)點(diǎn)間通過消息傳遞的方式搜尋和確定聚類中心,主要傳遞2種類型的消息,即吸引度(responsibility,)和歸屬度(availability,)。(,)表示從點(diǎn)發(fā)送到候選聚類中心的數(shù)值消息,它反映了適合作為的類代表點(diǎn)所積累的證據(jù);(,)表示從候選聚類中心發(fā)送到的數(shù)值消息,反映了選擇作為其類代表點(diǎn)的合適程度所積累的證據(jù)。(,)和(,)越大,則點(diǎn)為聚類中心的可能性越大,并且點(diǎn)隸屬于以點(diǎn)為聚類中心的聚類可能性也越大[24-26]。圖 1為消息的傳遞過程。

      圖1 消息傳遞過程

      AP聚類算法依照式(2)和式(3)分別迭代更新吸引度(,)和歸屬度(,)的值。

      其中

      其中

      式中為時刻;為阻尼系數(shù),取值[0.5, 1],一般取0.9,用于保證算法的收斂[27-29]。(,)的初始值為0。

      AP聚類算法在每次迭代后將)+(,)>0的數(shù)據(jù)點(diǎn)作為簇中心。當(dāng)?shù)螖?shù)超過設(shè)置閾值時或者當(dāng)聚類中心連續(xù)多次迭代不發(fā)生改變時終止迭代。確定所有的聚類中心后,將其余的數(shù)據(jù)點(diǎn)分配到相應(yīng)的類中心[30-31]。具體的算法流程如圖2所示。

      圖2 AP聚類算法流程

      Step2:由式(2)和式(3)計(jì)算樣本點(diǎn)間的吸引度和歸屬度值,即()和(,)。

      Step3:迭代更新()和(,),每次迭代更新后,將)+(,)>0的數(shù)據(jù)對象選作為簇中心。

      Step4:當(dāng)?shù)螖?shù)超過最大迭代次數(shù)時(如maxits為1 000次)或者當(dāng)聚類中心連續(xù)多少次迭代不發(fā)生改變時終止迭代(如convits為100次),確定類中心及各類的樣本點(diǎn),否則返回Step2,繼續(xù)計(jì)算。

      Step5:將剩余的數(shù)據(jù)點(diǎn)根據(jù)相似度劃分到各個類當(dāng)中,執(zhí)行完畢,算法結(jié)束。

      2 傳感器優(yōu)化布局試驗(yàn)及結(jié)果分析

      2.1 試 驗(yàn)

      試驗(yàn)場地位于安徽農(nóng)業(yè)大學(xué)國家高新技術(shù)農(nóng)業(yè)園(117.210°E,31.937°N,海拔29 m),選取約16 000 m2的茶園,橫向約190 m、縱向約84 m。本試驗(yàn)采用無線傳感網(wǎng)絡(luò)傳輸數(shù)據(jù),經(jīng)測試無線傳感網(wǎng)的可靠傳輸距離約40 m,為保證無線傳感網(wǎng)路的全覆蓋,選擇橫向約38 m、縱向約14 m的間隔布置傳感器,橫向布置5個傳感器,縱向布置5個傳感器,共計(jì)25個傳感器檢測點(diǎn),各數(shù)據(jù)采集點(diǎn)分別用A1、A2、A3、…、E4、E5進(jìn)行標(biāo)記。傳感器檢測點(diǎn)分布如圖3所示。

      注:A1-E5為數(shù)據(jù)采集點(diǎn)。

      傳感器選用石家莊雷光電子科技有限公司的SWR-100W土壤墑情傳感器,該傳感器基于頻域反射原理測量土壤的質(zhì)量含水率。各檢測點(diǎn)均采集土壤表面之下25 cm土壤墑情數(shù)據(jù)。

      實(shí)時采集茶園的土壤含水率和相對含水率2個參數(shù)。土壤含水率是指土壤中水分與烘干土質(zhì)量的比值,%;土壤相對含水率為土壤含水率占田間持水量的百分比,%;田間持水量是指毛管懸著水達(dá)到最大時的土壤含水率。

      2.2 結(jié)果與分析

      2018年7月28日、7月31日和8月3日分別采集了試驗(yàn)區(qū)域25 cm深度的土壤墑情數(shù)據(jù)。7月28日有雷陣雨,28~36 ℃,7月31日晴天,27~35 ℃,8 月3日多云,26~34 ℃。對采集的數(shù)據(jù)運(yùn)用Kriging最優(yōu)內(nèi)插估值方法繪制試驗(yàn)區(qū)的土壤墑情空間分布如圖4所示。

      圖4 茶園土壤含水率及相對含水率空間分布

      由圖4可以看出,試驗(yàn)區(qū)土壤含水率分布變化總體呈西南-東北向遞增趨勢,其相對含水率的最大變化范圍約15%,土壤墑情分布存在一定的差異。試驗(yàn)區(qū)的西邊空間分布呈現(xiàn)平緩均勻的特征,中間區(qū)域出現(xiàn)凸點(diǎn)。

      試驗(yàn)場土壤為下蜀系黃棕壤,土層較厚、質(zhì)地黏重。對各采集點(diǎn)用環(huán)刀采集25 cm深度的土壤,并用鋁合法進(jìn)測試土壤的孔隙度,孔隙度平均值為47.37%,標(biāo)準(zhǔn)差1.9%,試驗(yàn)場的土壤性質(zhì)具有較好的均一性。試驗(yàn)茶園的西南方向地勢較高,東及偏北方向地勢偏低,約有5°的坡度,場地的中間有一片地勢低洼的區(qū)域,與空間分布圖中的凸點(diǎn)位置重合,該區(qū)域位于C3和C4采集點(diǎn)附近。對照圖4土壤墑情的空間分布,初步判斷試驗(yàn)區(qū)的土壤墑情分布的差異性主要受場地的地形結(jié)構(gòu)影響。

      在AP聚類計(jì)算過程中,設(shè)定=0.9,最大迭代次數(shù)為1 000,當(dāng)聚類中心連續(xù)迭代100次不發(fā)生改變時終止迭代(即convits為100次)。偏向參數(shù)為相似度矩陣對角線上的中值,7月28日、7月31日和8月3日數(shù)據(jù)值分別為-86.10、-120.63、-102.44。設(shè)置3 d的的倍率值一致,即3、5、10、15、20和25。結(jié)合試驗(yàn)區(qū)域土壤墑情數(shù)據(jù)進(jìn)行AP聚類的迭代計(jì)算,圖5為7月28日土壤含水率和相對含水率在不同值下的聚類結(jié)果。表 1為3 d不同值的聚類數(shù)和聚類中心分布情況。

      由圖5和表1可知,當(dāng)選取10、15、20和25的聚類參數(shù)時,7月28日的聚類數(shù)為2,不同值聚類中心都是A2、D4采集點(diǎn),聚類結(jié)果具有較好的一致性。同樣,7月31日和8月3日的聚類結(jié)果與7月28日相似,選取10、15、20和25的參數(shù)時聚類結(jié)果也為2,分別為A1、C4和A2、D5采集點(diǎn),當(dāng)天的聚類結(jié)果也具備較好的一致性??梢姡x擇10、15、20和25的聚類參數(shù)對試驗(yàn)區(qū)域的土壤墑情數(shù)據(jù)進(jìn)行聚類,能夠得到穩(wěn)定和一致的結(jié)果,3 d的聚類結(jié)果均為2。因此,通過AP聚類算法對試驗(yàn)區(qū)域土壤墑情傳感器進(jìn)行優(yōu)化布置,傳感器數(shù)量可由25個減少為2個。

      注:p為AP聚類算法的偏向參數(shù)。下同。

      表1 3 d試驗(yàn)數(shù)據(jù)對應(yīng)不同p值的聚類結(jié)果

      分別將3 d試驗(yàn)各聚類中心的相對含水率與類均值進(jìn)行比較,如表2所示。7月28日A1采集點(diǎn)的土壤相對含水率與類均值相差-1.15%,絕對值大于A2與類均值的相對偏差(-0.58%),D4采集點(diǎn)的土壤相對含水率與類均值相差1.85%,絕對值小于C4和D5與類均值的相對偏差(2.09%~2.31%)。另外2 d試驗(yàn)結(jié)果趨勢一致:A1與A2相比,3 d試驗(yàn)與類均值的相對偏差(絕對值)均較小,C4、D4和D5相比,D4與與類均值的相對偏差(絕對值)均較小。因此,A2和D4采集點(diǎn)的相對含水率與試驗(yàn)區(qū)平均值接近,能夠代表試驗(yàn)區(qū)域的土壤墑情,優(yōu)化布局點(diǎn)選擇在A2、D4點(diǎn)布置傳感器較為合理。

      表2 3 d試驗(yàn)的類中心實(shí)測值與平均相對含水率

      為驗(yàn)證A2、D4點(diǎn)的代表性,于2019年1月18日采集了這2點(diǎn)的含水率,另外在試驗(yàn)區(qū)另外隨機(jī)選取了13個位置(如圖6所示)采集數(shù)據(jù),計(jì)算土壤相對含水率。結(jié)果表明,A2和D4點(diǎn)的土壤相對含水率分別為27.9%和37%,其平均值為32.45%,13個采集點(diǎn)土壤相對含水率的平均值為32.7%,2個平均值數(shù)值接近,相對偏差較小,為0.76%??梢?,在A2、D4采集點(diǎn)布置傳感器能夠代表整個試驗(yàn)區(qū)域的土壤墑情信息。綜上,本文試驗(yàn)區(qū)茶園選擇分別位于A2和D4的 2個點(diǎn)布置土壤墑情傳感器,測定的值能夠反映茶園土壤墑情的整體狀況。

      注:1~13為隨機(jī)布設(shè)的采樣點(diǎn)。

      3 總 結(jié)

      本文提出了基于AP聚類算法的土壤墑情傳感器布局策略,并將其應(yīng)用于茶園土壤墑情的傳感器布局優(yōu)化問題。在保證茶園傳感網(wǎng)絡(luò)全覆蓋的基礎(chǔ)上,實(shí)時采集各節(jié)點(diǎn)的土壤墑情數(shù)據(jù)。采用本文的優(yōu)化方案,試驗(yàn)區(qū)傳感器數(shù)量從25個減少至2個,顯著降低了系統(tǒng)成本,降低了數(shù)據(jù)的冗余度。為驗(yàn)證結(jié)果,在試驗(yàn)區(qū)隨機(jī)采集了土壤墑情數(shù)據(jù),計(jì)算了土壤平均相對含水率。經(jīng)與聚類中心的數(shù)據(jù)進(jìn)行比較,平均相對含水率與聚類中心值接近,相對偏差為0.76%,表明聚類中心的采集數(shù)據(jù)能夠代表試驗(yàn)區(qū)域的土壤墑情。

      采用的AP聚類算法通用性強(qiáng)、穩(wěn)定可靠、對初值不敏感,易于與其他算法進(jìn)行融合。雖然本文的研究對象是茶園,但該方法對于解決各類農(nóng)業(yè)節(jié)水灌溉系統(tǒng)的傳感器優(yōu)化問題可提供有益的借鑒。本文的試驗(yàn)選取3 d時間進(jìn)行試驗(yàn),聚類結(jié)果具有一致性和穩(wěn)定性,排除了一定的偶然性,但尚有一定的局限性,沒有考慮不同氣象條件下的聚類情況,后期將選擇不同氣象條件下以及不同特性的農(nóng)田開展試驗(yàn),提升研究結(jié)果的通用性。

      [1] Cheng Shunfeng, Michael H Azarian, Michael G Pecht. Sensor system for prognostics and health management[J]. Sensors, 2010, 10(6): 5774-5797.

      [2] 朱喜華,李穎暉,李寧,等. 基于改進(jìn)離散粒子群算法的傳感器布局優(yōu)化設(shè)計(jì)[J]. 電子學(xué)報(bào),2013,41(10):2104-2108. Zhu Xihua, Li Yinhui, Li Ning, et al. Optimal sensor placement design based on improved discrete PSO algorithm[J]. Acta Electronica Sinica, 2013, 41(10): 2104-2108. (in Chinese with English abstract)

      [3] 陳仲生,楊擁民,李聰,等. 基于振動傳遞符號有向圖的齒輪箱嵌入式傳感器優(yōu)化配置模型與算法[J]. 航空動力學(xué)報(bào),2009,24(10):2384-2390. Chen Zhongsheng, Yang Yongmin, Li Cong, et al. Model and algorithm of optimal embedded sensor placement for gearboxes based on signed directed graph of vibration progagation[J]. Journal of Aerospace Power, 2009, 24(10): 2384-2390. (in Chinese with English abstract)

      [4] 王宏力,張忠泉,崔祥祥,等. 基于改進(jìn)PSO算法的實(shí)時故障監(jiān)測診斷測試集優(yōu)化[J]. 系統(tǒng)工程與電子技術(shù),2011,33(4):958-962. Wang Hongli, Zhang Zhongquan, Cui Xiangxiang, et al. Test optimization of real-time monitoring and fault diagnosis system based on improved particle swarm optimization[J]. Systems Engineering and Electonics, 2011, 33(4): 958-962. (in Chinese with English abstract)

      [5] 楊帆,蕭德云. 故障檢測的可靠性描述及傳感器分布優(yōu)化算法[J]. 應(yīng)用科學(xué)學(xué)報(bào),2006,24(2):125-130. Yang Fan, Xiao Deyun. Reliability description of fault detection and optimization algorithm of sensor location[J]. Journal of Applied Sciences, 2006, 24(2): 125-130. (in Chinese with English abstract)

      [6] 劉麗萍. 無線傳感器網(wǎng)絡(luò)節(jié)能覆蓋[D]. 杭州:浙江大學(xué),2006. Liu Liping. Energy-efficient Coverage in Wireless Sensor Networks[D]. Hangzhou: Zhejiang University, 2006. (in Chinese with English abstract)

      [7] Lin Frank Y S, Chiu P L. A simulated annealing algorithm for energy-efficient sensor network design[C]//International Symposium on Modeling & Optimization in Mobile. 2005.

      [8] Wang Y C, Hu C C, Tseng Y C. Efficient placement and dispatch of sensors in a wireless sensor network[J]. IEEE Transactions on Mobile Computing, 2007, 7(2):262-274.

      [9] 趙偉霞, 李久生, 王珍, 等. 滴灌均勻性對土壤水分傳感器埋設(shè)位置的影響[J]. 農(nóng)業(yè)工程學(xué)報(bào),2018,34(9):123-129. Zhao Weixia, Li Jiusheng, Wang Zhen, et al. Influence of drip irrigation uniformity on buried position of soil moisture sensor[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(9): 123-129.(in Chinese with English abstract)

      [10] 蔣杰,方力,張鶴穎. 無線傳感器網(wǎng)絡(luò)最小連通覆蓋問題求解算法[J]. 軟件學(xué)報(bào) 2006,17(2):175-184. Jiang Jie, Fang Li, Zhang Heying. An algorithm for minimal connected cover set problem in wireless sensor networks. Journal of Software, 2006,17(2):175?184. (in Chinese with English abstract)

      [11] 李飚,魏正英,張育斌,等. 土壤墑情傳感器布點(diǎn)算法研究[J]. 節(jié)水灌溉,2015(2):72-76. Li Biao, Wei Zhengying, Zhang Yubin, et al. Optimal placement algorithm of soil moisture sensors [J]. Water-saving Irrigation, 2015(2):72-76. (in Chinese with English abstract)

      [12] 姚月峰,滿秀玲. 毛烏素沙地不同林齡沙柳表層土壤水封空間異質(zhì)性[J]. 水土保持學(xué)報(bào),2007,21(1):112-115.Yao Yuefeng, Man Xiuling. Spatial Heterogeneity of top soil moisture in different stand age of salix psammophila in Mu US sandy land [J]. Journal of Soil and Water Conservation, 2007, 21(1): 112-115. (in Chinese with English abstract)

      [13] 黃弈龍,陳利頂,傅伯杰,等. 黃土丘陵小流域土壤水分空間格局及其影響因素[J]. 自然資源學(xué)報(bào),2005,20(4):483-492. Huang Yilong, Chen Liding, Fu Bojie, et al. Spatial pattern of soil moisture and its influencing factors in small watershed of loess hilly region [J]. Journal of Natural Resources, 2005, 20(4): 483-492. (in Chinese with English abstract)

      [14] 潘顏霞,王新平,蘇延桂,等. 不同植被類型沙地表層土壤水分變化特征[J]. 水土保持學(xué)報(bào),2007,21(5):107-109. Pan Yanxia, Wang Xinping, Su Yangui, et al. Variability characteristics of surface soil moisture content in sand areas covered by different vegetation types [J]. Journal of Soil and Water Conservation, 2007, 21(5): 107-109. (in Chinese with English abstract)

      [15] 趙永宏,劉賢德,張學(xué)龍,等. 祁連山區(qū)亞高山灌叢土壤含水量的空間分布與月份變化規(guī)律[J]. 自然資源學(xué)報(bào),2016,31(4):672-681. Zhao Yonghong, Liu Xiande, Zhang Xuelong, et al. Spatial distribution and monthly variation of soil water content in subalpine shrubs in Qilian Mountains [J]. Journal of Natural Resources, 2016, 31(4): 672-681. (in Chinese with English abstract)

      [16] 許迪,Schmicl R. 田間土壤特性的空間相關(guān)結(jié)構(gòu)分析及其分布描述[J]. 灌溉排水,1996,15(4):16-20. Xu Di, Schmicl R. Spatial correlation structure analysis and distribution description of field soil properties[J]. Irrigation and Drainage, 1996, 15(4):16-20 (in Chinese with English abstract)

      [17] 劉自豪,張斌,祝寧,等. 基于改進(jìn)AP聚類算法的自學(xué)習(xí)應(yīng)用層DDoS檢測方法[J]. 計(jì)算機(jī)研究與發(fā)展,2018,55(6):1236-1246.Liu Zihao, Zhang Bin, Zhu Ning, et al. Adaptive app-ADDoS detection method based on improved AP algorithm [J]. Computer research and development, 2018, 55(6): 1236-1246. (in Chinese with English abstract)

      [18] 唐丹,張正軍. 近鄰傳播聚類算法的優(yōu)化[J]. 計(jì)算機(jī)應(yīng)用,2017,37(z1):258-261.Tang Dan, Zhang Zhengjun. New Optimized affinity propagation clustering Algorithms [J]. Computer Applications, 2017, 37 (z1): 258-261. (in Chinese with English abstract)

      [19] Frey B J, Dueck D. Clustering by passing messages between data points[J]. Science, 2007, 315(5814): 972-976.

      [20] 董俊,王鎖萍,熊范綸.可變相似性度量的近鄰傳播聚類[J]. 電子與信息學(xué)報(bào),2010,32(3):509-514. Dong Jun, Wang Suoping, Xiong Fanlun. Affinity Propagation clustering based on variable-similarity measure[J]. Journal of Electronics and Information, 2010, 32(3): 509-514. (in Chinese with English abstract)

      [21] 王開軍,張軍英,李丹,等. 自適應(yīng)仿射傳播聚類[J]. 自動化學(xué)報(bào),2007,33(12):1242-1246. Wang Kaijun, Zhang Junying, Li Dan, et al. Adaptive affinity propagation clustering [J]. Journal of Automation, 2007, 33(12): 1242-1246. (in Chinese with English abstract)

      [22] Wei Z, Wang Y, He S, et al. A novel intelligent method for bearing fault diagnosis based on affinity propagation clustering and adaptive feature selection[J]. Knowledge-Based Systems, 2017, 116:1-12.

      [23] 孫勁光,趙欣. 一種改進(jìn)近鄰傳播聚類的圖像分割算法[J]. 計(jì)算機(jī)工程與應(yīng)用,2017,53(6):178-182. Sun Jinguang, Zhao Xin. Image segmentation algorithm based on improved affinity propagation clustering.Computer Engineering and Applications, 2017, 53(6):178-182. (in Chinese with English abstract)

      [24] 儲岳中,劉恒,張學(xué)鋒. 基于遷移AP聚類與稀疏表示的遙感圖像分類[J]. 計(jì)算機(jī)工程與設(shè)計(jì),2018,39(2):547-550. Chu Yuezhong, Liu Heng, Zhang Xuefeng. Remote sensing image classification method based on tansfer affinity propagation and sparse representation[J]. Computer Engineering and Design, 2018, 39(2): 547-550. (in Chinese with English abstract)

      [25] 江頡,王卓芳,陳鐵明,等. 自適應(yīng)AP聚類算法及其在入侵檢測中的應(yīng)用[J]. 通信學(xué)報(bào),2015,36(11):118-126. Jiang Jie, Wang Zhuofang, Chen Tieming, et al.Adaptive AP clustering algorithm and its application on intrusion detection [J]. Journal of Communications, 2015, 36(11): 118-126. (in Chinese with English abstract)

      [26] 趙延龍,滑楠. 基于初始偏向度的AP算法聚類性能優(yōu)化研究[J]. 計(jì)算機(jī)應(yīng)用研究,2018(2):372-374. Zhao Yanlong, Hua Nan. Research on optimization of AP algorithm performance based on initial preference[J]. Computer Application, 2018(2): 372-374. (in Chinese with English abstract)

      [27] 陳瑩. 基于AP聚類算法的圖像分割技術(shù)研究[D]. 吉林:吉林建筑大學(xué),2015. Chen Ying. Research on Image Segmentation Technology Based on AP Clustering Algorithm[D]. Jilin: Jilin University of Architecture, 2015. (in Chinese with English abstract)

      [28] Yang C, Liu S, Bruzzone L, et al. A feature-metric-based affinity propagation technique for feature selection in hyperspectral image classification[J]. IEEE Geoscience & Remote Sensing Letters, 2013, 10(5): 1152-1156.

      [29] Gao J, Ma Z, Qin Y, et al. Application of affinity propagation clustering algorithm in fault diagnosis of metro vehicle auxiliary inverter[J]. Lecture Notes in Electrical Engineering, 2014, 288:3-9.

      [30] 劉曉勇,付輝. 一種快速AP聚類算法[J]. 山東大學(xué)學(xué)報(bào):工學(xué)版,2011,41(4):20-23. Liu Xiaoyong, Fu Hui. A fast affinity propagation clustering algorithm[J]. Journal of Shandong University: Engineering Edition, 2011, 41(4):20-23. (in Chinese with English abstract)

      [31] 肖宇,于劍. 基于近鄰傳播算法的半監(jiān)督聚類[J]. 軟件學(xué)報(bào),2008,19(11):2803-2813. Xiao Yu, Yu Jian. Semi-supervised clustering based on affinity propagation algorithms[J]. Journal of Software, 2008, 19(11): 2803-2813. (in Chinese with English abstract)

      Layout optimization of soil moisture sensor in tea plantation based on affinity propagation clustering algorithm

      Zhang Wu1, Zhang Manman1, Hong Xun1, Jiang Zhaohui1※, Jiang Yuelin2

      (1.,,230036,; 2.,,230036,)

      Aiming at the layout problem of soil moisture sensors for water-saving irrigation, we proposed an optimal layout strategy of soil moisture sensors based on affinity propagation (AP) clustering algorithm. The soil moisture of tea plantation was as the research object. The tea plantation had 84-m width and 190-m length. Following the conventional method, 25 sensor nodes were evenly arranged in rectangular mode in tea plantation experimental area in order to guarantee full coverage of tea plantation sensor network. Soil moisture data of each sensor node in the test area was collected in real time for 3 days. The optimization of sensors was conducted based on soil water content and relative water content by AP clustering algorithm. Different clustering parameters were selected. The AP clustering algorithm was used to construct similarity matrix of node soil water content, to iteratively calculate the responsibility and availability of each node, and to form the clustering number and clustering center. When the clustering parameters were 10, 15, 20 and 25 times of preference, the AP clustering algorithm was used to calculate the soil moisture data in the experimental area for 3 days, the stable and consistent clustering results were obtained. Results showed that soil water content in the tested plantation presented an increasing trend from southwest to northeast and the largest difference of relative water content was 15%. The change is related to the topography of the tested area. For AP clustering, the maximum iterative times was designed as 1 000. Based on the results, the clustering result in the 3 days was 2. The number of sensors optimized by AP clustering algorithm was reduced from 25 to 2. The class mean of the relative water content of the soil in the experimental area was calculated, and compared with the relative water content of soil in the collection points of the cluster center, and the relative bias between them was less than 5%. The relative water content of the collection points in the cluster center was close to the average value of the experimental area, which indicated that the data collected by the cluster center can represent soil moisture situation in the experimental tea plantation. In order to verify the validity of this method, soil moisture data were collected randomly at 13 locations in the experimental area on January 2019. Results showed that the soil average relative water content of tea plantation in the experimental area based on 13 sampling points was 32.7%, the relative water content of soil in the cluster center based on 2 sensors was respectively 27.9% and 37% with an average in the cluster center of 32.45%. Compared with the average relative moisture in the experimental area, the relative bias was only 0.76%. It means that the AP clustering algorithm can optimize the distribution of soil moisture sensors in the experimental tea plantation. The relative soil moisture collected by the cluster center could reflect the overall situation of soil moisture in the tea plantation as long as using only 2 sensors arranged in the cluster center node determined by the optimization calculation. Thus, the AP clustering algorithm is suggested to use in optimization of the sensor layout, which can reduce the redundancy of data and accordingly realize cost saving in agricultural production system.

      soil moisture; sensors; clustering algorithms; layout optimization; affinity propagation clustering

      2018-09-16

      2019-02-15

      2018年安徽省重點(diǎn)研究和開發(fā)計(jì)劃項(xiàng)目(1804a07020108);2017年安徽省科技重大專項(xiàng)計(jì)劃(17030701049);2016年農(nóng)業(yè)部農(nóng)業(yè)物聯(lián)網(wǎng)技術(shù)集成與應(yīng)用重點(diǎn)實(shí)驗(yàn)室開放基金(2016KL05)

      張武,副教授,博士,主要從事計(jì)算機(jī)控制及農(nóng)業(yè)物聯(lián)網(wǎng)研究。Email:zhangwu@ahau.edu.cn

      江朝暉,教授,博士,主要從事農(nóng)業(yè)信息檢測與處理研究。Email:jiangzh@ahau.edu.cn

      10.11975/j.issn.1002-6819.2019.06.013

      S126

      A

      1002-6819(2019)-06-0107-07

      張 武,張嫚嫚,洪 汛,江朝暉,蔣躍林. 基于近鄰傳播算法的茶園土壤墑情傳感器布局優(yōu)化[J]. 農(nóng)業(yè)工程學(xué)報(bào),2019,35(6):107-113. doi:10.11975/j.issn.1002-6819.2019.06.013 http://www.tcsae.org

      Zhang Wu, Zhang Manman, Hong Xun, Jiang Zhaohui, Jiang Yuelin. Layout optimization of soil moisture sensor in tea plantation based on affinity propagation clustering algorithm[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2019, 35(6): 107-113. (in Chinese with English abstract) doi:10.11975/j.issn.1002-6819.2019.06.013 http://www.tcsae.org

      猜你喜歡
      墑情布局含水率
      昆明森林可燃物燃燒機(jī)理研究
      墑情會商,助力備耕春播
      基于表面能的濕煤顆粒含水率表征方法
      土壤墑情監(jiān)測的意義與技術(shù)規(guī)范
      弱膨脹土增濕變形量試驗(yàn)及路堤填筑分析
      BP的可再生能源布局
      能源(2017年5期)2017-07-06 09:25:57
      原油含水率在線測量技術(shù)研究
      電子制作(2016年1期)2016-11-07 08:42:56
      VR布局
      2015 我們這樣布局在探索中尋找突破
      開封市土壤墑情監(jiān)測系統(tǒng)建設(shè)研究
      凭祥市| 兴仁县| 孝昌县| 枣强县| 芜湖县| 荆门市| 冕宁县| 且末县| 曲麻莱县| 周宁县| 罗定市| 昂仁县| 乡宁县| 昆山市| 陆河县| 三台县| 德昌县| 泰宁县| 嘉祥县| 东莞市| 永福县| 开江县| 道孚县| 丹江口市| 三台县| 滨海县| 友谊县| 景谷| 罗定市| 客服| 盐亭县| 耿马| 荆门市| 永平县| 新野县| 宁晋县| 临桂县| 碌曲县| 新干县| 东乡族自治县| 桦南县|