国产黄在线观看免费观看不卡_日韩精品无码一区二区三区_久久人人爽天天玩人人妻精品_漂亮人妻去按摩被按中出

改(gai)革(ge)(ge)開(kai)(kai)放以來，我(wo)國城(cheng)(cheng)市(shi)供(gong)(gong)水(shui)(shui)(shui)能力和供(gong)(gong)水(shui)(shui)(shui)質(zhi)量(liang)不斷提高，現已建(jian)成規模龐(pang)大的(de)(de)供(gong)(gong)水(shui)(shui)(shui)設(she)施，根據《中國城(cheng)(cheng)市(shi)建(jian)設(she)統計年(nian)鑒（2017年(nian)）》數據顯示，全(quan)(quan)國已建(jian)成城(cheng)(cheng)市(shi)公(gong)共供(gong)(gong)水(shui)(shui)(shui)廠(chang)約2 880個(ge)。近年(nian)來，我(wo)國各(ge)地(di)按照相關政(zheng)策法規的(de)(de)要(yao)求(qiu)，開(kai)(kai)展(zhan)了城(cheng)(cheng)市(shi)供(gong)(gong)水(shui)(shui)(shui)水(shui)(shui)(shui)量(liang)、水(shui)(shui)(shui)質(zhi)、水(shui)(shui)(shui)壓等監(jian)(jian)測能力的(de)(de)建(jian)設(she)，具備了一定的(de)(de)供(gong)(gong)水(shui)(shui)(shui)安全(quan)(quan)監(jian)(jian)管能力。伴隨(sui)著在(zai)線監(jian)(jian)測手(shou)段和信息處理(li)技術的(de)(de)發(fa)展(zhan)，我(wo)國地(di)方政(zheng)府和城(cheng)(cheng)市(shi)供(gong)(gong)水(shui)(shui)(shui)單位對于革(ge)(ge)新供(gong)(gong)水(shui)(shui)(shui)行業的(de)(de)傳統監(jian)(jian)管方式(shi)、提高供(gong)(gong)水(shui)(shui)(shui)監(jian)(jian)管的(de)(de)自(zi)動化(hua)和智(zhi)能化(hua)水(shui)(shui)(shui)平，表(biao)現出了濃厚(hou)興趣(qu)，“智(zhi)慧水(shui)(shui)(shui)務”建(jian)設(she)在(zai)各(ge)地(di)蓬勃開(kai)(kai)展(zhan)。

“十(shi)(shi)一五”和(he)“十(shi)(shi)二五”期間，通過水(shui)(shui)(shui)(shui)(shui)(shui)專項課題的(de)開展，初步建(jian)立(li)了國(guo)(guo)家(jia)、省、市(shi)(shi)三級的(de)城(cheng)市(shi)(shi)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)監測(ce)(ce)(ce)預(yu)警(jing)系統(tong)技術平臺(tai)。但一方面(mian)當前(qian)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)監管中(zhong)存在(zai)(zai)實測(ce)(ce)(ce)指(zhi)標(biao)不(bu)全(quan)面(mian)、監測(ce)(ce)(ce)頻率(lv)不(bu)達(da)標(biao)、風(feng)險預(yu)警(jing)不(bu)及(ji)時(shi)等問題，另一方面(mian)仍有部分地區的(de)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)運營管理相(xiang)關業(ye)務是以現場經驗判(pan)斷、手動操作、人工報數(shu)(shu)等傳統(tong)手段為主，大(da)中(zhong)城(cheng)市(shi)(shi)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)單(dan)位每日(ri)積(ji)累的(de)海量供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)數(shu)(shu)據(ju)所(suo)包含的(de)信(xin)息幾乎尚未(wei)被挖(wa)掘利用(yong)。據(ju)初步統(tong)計(ji)，我國(guo)(guo)直轄市(shi)(shi)、計(ji)劃單(dan)列市(shi)(shi)、省會(hui)城(cheng)市(shi)(shi)等36個(ge)(ge)重點城(cheng)市(shi)(shi)的(de)130多個(ge)(ge)公(gong)共供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)廠(chang)，每年積(ji)累的(de)水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)信(xin)息就多達(da)1 350萬余條。當前(qian)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)大(da)數(shu)(shu)據(ju)所(suo)蘊(yun)藏的(de)信(xin)息利用(yong)潛(qian)能已(yi)越來越為業(ye)內所(suo)共識，部分城(cheng)市(shi)(shi)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)單(dan)位開始利用(yong)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)大(da)數(shu)(shu)據(ju)開展了漏(lou)損控制(zhi)、管網健(jian)康(kang)度評價管理等工作。國(guo)(guo)外一些(xie)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)單(dan)位對(dui)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)數(shu)(shu)據(ju)的(de)應用(yong)也進(jin)行了一些(xie)探索，例如英國(guo)(guo)聯合水(shui)(shui)(shui)(shui)(shui)(shui)務用(yong)供(gong)(gong)(gong)(gong)水(shui)(shui)(shui)(shui)(shui)(shui)數(shu)(shu)據(ju)預(yu)測(ce)(ce)(ce)未(wei)來用(yong)水(shui)(shui)(shui)(shui)(shui)(shui)量，從而提前(qian)做好生(sheng)產(chan)預(yu)案(an)；荷蘭的(de)Vitens公(gong)司針對(dui)管網建(jian)立(li)了預(yu)警(jing)系統(tong)可以在(zai)(zai)2 min內識別爆(bao)管事件；歐(ou)盟資助的(de)智慧水(shui)(shui)(shui)(shui)(shui)(shui)務項目也在(zai)(zai)研究(jiu)根據(ju)用(yong)戶用(yong)水(shui)(shui)(shui)(shui)(shui)(shui)習慣預(yu)測(ce)(ce)(ce)未(wei)來用(yong)水(shui)(shui)(shui)(shui)(shui)(shui)量和(he)消費趨勢(shi)等。

當前我(wo)國各(ge)地(di)的供水(shui)(shui)監(jian)管(guan)信息化水(shui)(shui)平(ping)雖有(you)差(cha)異，但信息化建設進(jin)程正在(zai)(zai)提(ti)(ti)速發(fa)展。“十三(san)五”期間(jian)，依(yi)托“城市供水(shui)(shui)全過程監(jian)管(guan)平(ping)臺整合及(ji)業(ye)務化運行”課題，多家供水(shui)(shui)單位在(zai)(zai)建成了業(ye)務應(ying)用系統(tong)模塊基礎(chu)上，正在(zai)(zai)探(tan)索進(jin)一步提(ti)(ti)高信息應(ying)用效率、提(ti)(ti)高監(jian)管(guan)水(shui)(shui)平(ping)。

2.1 內部來源

2.1.1統計報表數據

一是可從(cong)城市供(gong)(gong)水(shui)(shui)(shui)(shui)單位(wei)、水(shui)(shui)(shui)(shui)質檢(jian)測(ce)(ce)機(ji)構等單位(wei)獲(huo)取(qu)水(shui)(shui)(shui)(shui)質數(shu)(shu)據(ju)(ju)，包括水(shui)(shui)(shui)(shui)源(yuan)水(shui)(shui)(shui)(shui)、水(shui)(shui)(shui)(shui)廠各工(gong)藝段進(jin)出(chu)水(shui)(shui)(shui)(shui)、出(chu)廠水(shui)(shui)(shui)(shui)、管網水(shui)(shui)(shui)(shui)、二(er)次供(gong)(gong)水(shui)(shui)(shui)(shui)、龍頭水(shui)(shui)(shui)(shui)等環節的水(shui)(shui)(shui)(shui)質檢(jian)測(ce)(ce)數(shu)(shu)據(ju)(ju)；二(er)是可從(cong)城市供(gong)(gong)水(shui)(shui)(shui)(shui)單位(wei)獲(huo)取(qu)生產數(shu)(shu)據(ju)(ju)，包括設施資產、設備工(gong)況、材料(liao)庫存(cun)、售水(shui)(shui)(shui)(shui)情(qing)況、供(gong)(gong)水(shui)(shui)(shui)(shui)用戶(hu)信(xin)(xin)息、供(gong)(gong)水(shui)(shui)(shui)(shui)管網信(xin)(xin)息、設備維護(hu)檢(jian)修記錄(lu)、服(fu)務(wu)投(tou)訴信(xin)(xin)息等。

2.1.2 設備自動監測數據

一是可(ke)(ke)從(cong)城市(shi)供水(shui)(shui)(shui)主管部門和城市(shi)供水(shui)(shui)(shui)單位(wei)(wei)收集獲取設備自動監測數據，包括在線采集的水(shui)(shui)(shui)量(liang)、水(shui)(shui)(shui)位(wei)(wei)、水(shui)(shui)(shui)質等實(shi)時(shi)數據；二是可(ke)(ke)從(cong)城市(shi)供水(shui)(shui)(shui)單位(wei)(wei)獲取現場作業數據，包括員(yuan)工通過移動設備人(ren)為實(shi)時(shi)遠傳的地(di)理(li)位(wei)(wei)置、供水(shui)(shui)(shui)用戶水(shui)(shui)(shui)量(liang)、事故特征、現場照片、視(shi)頻等數據。

2.2 外部來源

除(chu)了(le)獲取城市(shi)供水(shui)系統內各(ge)部(bu)門(men)(men)數據，還可從環保、水(shui)利、氣(qi)象、衛生健康等相(xiang)關部(bu)門(men)(men)獲取與城市(shi)供水(shui)相(xiang)關的水(shui)質、水(shui)文、氣(qi)象等數據。

此(ci)外，在不影響被訪問的網(wang)站正(zheng)常運行的前提下，采用符合(he)法(fa)律(lv)、法(fa)規的方(fang)式，例如(ru)網(wang)絡(luo)爬蟲(chong)等技術，可獲取政府(fu)機構、企業等組(zu)織提供的與供水服(fu)務有關的免費開放數據(ju)(ju)(ju)，包括供水水質(zhi)信(xin)息(xi)公開數據(ju)(ju)(ju)、供水事(shi)故信(xin)息(xi)、人口數據(ju)(ju)(ju)、建筑信(xin)息(xi)數據(ju)(ju)(ju)等。

3.1 水源和水廠大數據應用

3.1.1 水質風險關鍵指標篩選

在(zai)(zai)水(shui)質(zhi)(zhi)日常監(jian)測(ce)、風(feng)險(xian)預警和管控(kong)過程中(zhong)，通過對水(shui)源、水(shui)廠，以及輸配(pei)水(shui)過程中(zhong)的(de)水(shui)質(zhi)(zhi)指標及其(qi)環境類指標進行相關(guan)性分析，找出不同水(shui)質(zhi)(zhi)指標之間(jian)、水(shui)質(zhi)(zhi)指標與其(qi)它環境類指標之間(jian)的(de)內(nei)在(zai)(zai)關(guan)聯(lian)性，可篩(shai)選出水(shui)質(zhi)(zhi)風(feng)險(xian)關(guan)鍵(jian)指標。

以篩選(xuan)可預(yu)測水(shui)(shui)(shui)源水(shui)(shui)(shui)體富營(ying)養(yang)化的水(shui)(shui)(shui)質(zhi)(zhi)(zhi)預(yu)警指標為(wei)例，山(shan)東省(sheng)城(cheng)市(shi)供排(pai)水(shui)(shui)(shui)水(shui)(shui)(shui)質(zhi)(zhi)(zhi)監測中心對某(mou)市(shi)水(shui)(shui)(shui)庫近5年的原(yuan)水(shui)(shui)(shui)水(shui)(shui)(shui)質(zhi)(zhi)(zhi)月檢數(shu)據進行了整理分析，包括溶解氧、總磷(lin)、總氮、氨(an)(an)(an)氮、硝酸鹽(yan)（以N計(ji)）、氮磷(lin)比(bi)、pH、渾(hun)濁度(du)、葉綠素a等9項(xiang)水(shui)(shui)(shui)質(zhi)(zhi)(zhi)指標。皮爾森相(xiang)關(guan)系數(shu)計(ji)算結果顯(xian)(xian)示，原(yuan)水(shui)(shui)(shui)中葉綠素a與硝酸鹽(yan)、總氮、pH、總磷(lin)、氮磷(lin)比(bi)、氨(an)(an)(an)氮濃度(du)存在(zai)顯(xian)(xian)著相(xiang)關(guan)性。進一步分析發現，氨(an)(an)(an)氮和磷(lin)元素是水(shui)(shui)(shui)源水(shui)(shui)(shui)中藻類(lei)增長(chang)最(zui)重要的限制因素。因此(ci)，初步篩選(xuan)將pH、總磷(lin)和氨(an)(an)(an)氮指標納(na)入(ru)預(yu)測水(shui)(shui)(shui)源水(shui)(shui)(shui)體富營(ying)養(yang)化趨(qu)勢(shi)的預(yu)警指標。

3.1.2 水質風險預警模型建立

以歷史水(shui)質(zhi)數(shu)據、相關(guan)水(shui)文及環境類等數(shu)據為基(ji)礎，通過應(ying)用各(ge)類數(shu)據特征(zheng)挖掘與(yu)分析技術，構(gou)建時間序列(lie)、回歸分析等風險評估模型，可(ke)對(dui)水(shui)質(zhi)指(zhi)標的(de)未來數(shu)值(zhi)(zhi)和風險進行預(yu)測(ce)，通過單點閾值(zhi)(zhi)、多點聯動等方式進行水(shui)質(zhi)風險預(yu)警。

以(yi)預測原水高錳酸(suan)鹽(yan)指數(shu)超標風(feng)險為(wei)例，山(shan)東省城市供排水水質監測中心為(wei)做好水質風(feng)險預警，根(gen)(gen)據2012年5月(yue)～2016年2月(yue)的(de)高錳酸(suan)鹽(yan)指數(shu)月(yue)檢數(shu)據，預測未來一(yi)段(duan)時間(jian)的(de)高錳酸(suan)鹽(yan)指數(shu)月(yue)度(du)平均值。根(gen)(gen)據數(shu)據波(bo)動特征(zheng)，選擇指數(shu)平滑(hua)法模型(xing)進(jin)行分析，預測結果顯示2016年3～8月(yue)的(de)高錳酸(suan)鹽(yan)指數(shu)月(yue)均濃度(du)不存在超標風(feng)險。

圖1 高錳酸鹽(yan)指數濃度預測曲(qu)線

3.1.3 水廠運行工藝調整輔助決策

通過分析原水(shui)(shui)(shui)(shui)(shui)(shui)(shui)關(guan)鍵水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質指標(biao)在工(gong)藝(yi)流程中的變化情(qing)況，并(bing)對工(gong)藝(yi)運行參數如藥耗(hao)、濾(lv)池反沖洗周期、排泥(ni)周期等，以(yi)及出(chu)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質情(qing)況同(tong)步分析，可基于不(bu)同(tong)進出(chu)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質條件下的運行工(gong)況和水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質預(yu)警(jing)結果(guo)構(gou)建工(gong)藝(yi)調整(zheng)輔(fu)助決策(ce)模型(xing)(xing)。當面臨水(shui)(shui)(shui)(shui)(shui)(shui)(shui)源地(di)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質突變、水(shui)(shui)(shui)(shui)(shui)(shui)(shui)廠藥耗(hao)增加等相關(guan)參數變化問題(ti)時，可將相關(guan)信(xin)息作(zuo)為輸入參數，利用輔(fu)助決策(ce)模型(xing)(xing)模擬出(chu)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)情(qing)況，從而避免了人(ren)為判斷的主觀性。此外(wai)，輔(fu)助決策(ce)模型(xing)(xing)也可預(yu)測出(chu)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質達標(biao)條件下對應(ying)的水(shui)(shui)(shui)(shui)(shui)(shui)(shui)源地(di)水(shui)(shui)(shui)(shui)(shui)(shui)(shui)質預(yu)警(jing)值及工(gong)藝(yi)藥耗(hao)最(zui)小值。

北(bei)京首創股份(fen)有限(xian)公司(si)在經營華北(bei)某(mou)水(shui)(shui)(shui)(shui)廠時(shi)，為(wei)了提(ti)前準(zhun)備(bei)工(gong)藝(yi)預(yu)案(an)，基于(yu)2014～2017年實測的(de)(de)(de)(de)進(jin)出(chu)水(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)數據和(he)記錄的(de)(de)(de)(de)運行工(gong)況(kuang)數據，建立(li)“水(shui)(shui)(shui)(shui)源地(di)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)/水(shui)(shui)(shui)(shui)量(liang)-水(shui)(shui)(shui)(shui)廠藥(yao)耗-出(chu)水(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)”在不(bu)同區間下(xia)一一對應(ying)的(de)(de)(de)(de)關聯(lian)性，并采用(yong)人工(gong)神(shen)經網絡技術建立(li)了工(gong)藝(yi)調(diao)整輔(fu)助決策模型。模型的(de)(de)(de)(de)輸入值(zhi)主要包括進(jin)水(shui)(shui)(shui)(shui)條件(jian)（溫度(du)(du)、pH）、水(shui)(shui)(shui)(shui)源地(di)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)（渾(hun)濁度(du)(du)、色度(du)(du)、高錳酸鉀指數、細菌總數）、進(jin)水(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)量(liang)及藥(yao)品投(tou)加量(liang)（聚合(he)氯化鋁投(tou)加量(liang)、加氯量(liang)），輸出(chu)值(zhi)為(wei)經凈水(shui)(shui)(shui)(shui)工(gong)藝(yi)處理后的(de)(de)(de)(de)對應(ying)出(chu)廠水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)。通過改變不(bu)同的(de)(de)(de)(de)工(gong)況(kuang)條件(jian)，可(ke)(ke)準(zhun)確快(kuai)捷地(di)預(yu)測得(de)到對應(ying)的(de)(de)(de)(de)出(chu)水(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)，并可(ke)(ke)同步計算所實現的(de)(de)(de)(de)污染(ran)物去除率，同時(shi)可(ke)(ke)反(fan)推(tui)在出(chu)水(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)達標(biao)要求(qiu)下(xia)，進(jin)水(shui)(shui)(shui)(shui)條件(jian)或各工(gong)況(kuang)工(gong)藝(yi)參(can)數的(de)(de)(de)(de)預(yu)警值(zhi)。

3.2 供水管網大數據應用

在明確(que)供(gong)水管(guan)網(wang)運(yun)行(xing)(xing)事故具(ju)體評(ping)(ping)價對象前提下，根據供(gong)水管(guan)網(wang)大數(shu)據，可(ke)建立模型對供(gong)水管(guan)網(wang)的(de)運(yun)行(xing)(xing)事故評(ping)(ping)價指(zhi)標發生概率(lv)進(jin)行(xing)(xing)定量預測。進(jin)一步結合管(guan)道級(ji)別(bie)、道路等(deng)級(ji)、人口密度(du)等(deng)因素，可(ke)通(tong)過(guo)構(gou)建定量判別(bie)指(zhi)標體系(xi)和評(ping)(ping)判標準，評(ping)(ping)估管(guan)道風險影響程度(du)，從而明確(que)管(guan)道修復/更(geng)新(xin)改造的(de)優先級(ji)，科學(xue)劃定供(gong)水管(guan)網(wang)修復/更(geng)新(xin)改造的(de)范圍。

以評(ping)估供水(shui)管(guan)網運(yun)行風(feng)(feng)險(xian)為(wei)(wei)例，深圳市水(shui)務（集(ji)團）有限(xian)公司選(xuan)擇爆管(guan)風(feng)(feng)險(xian)作為(wei)(wei)供水(shui)管(guan)網運(yun)行風(feng)(feng)險(xian)的(de)評(ping)估指(zhi)標，選(xuan)取管(guan)材、管(guan)徑(jing)、管(guan)齡(ling)、道路負荷、運(yun)行壓(ya)力、雜散(san)電(dian)流、是否發生破損(sun)等影響因子，采用隨機森林模(mo)型構建了(le)供水(shui)管(guan)網爆管(guan)風(feng)(feng)險(xian)評(ping)估模(mo)型，取得了(le)較好(hao)的(de)預測結果(guo)（見圖2），以該數(shu)值的(de)大小來(lai)量化評(ping)價供水(shui)管(guan)網運(yun)行風(feng)(feng)險(xian)，并作為(wei)(wei)制定供水(shui)管(guan)網更新改造計劃的(de)重要(yao)數(shu)據參考。

圖2 供(gong)水管網(wang)爆管概率(lv)預(yu)測結果(guo)與實際破損(sun)情(qing)況對比

利(li)用供水大數(shu)據開展(zhan)管(guan)網(wang)漏損控(kong)(kong)制(zhi)也是當(dang)前的(de)熱(re)點應用之一(yi)。伴隨著住(zhu)建(jian)部《城鎮(zhen)供水管(guan)網(wang)分區計(ji)量(liang)管(guan)理(li)工作指南——供水管(guan)網(wang)漏損管(guan)控(kong)(kong)體系構(gou)建(jian)（試行）》等相(xiang)關政(zheng)策文件的(de)出(chu)臺，基于分區計(ji)量(liang)管(guan)理(li)的(de)漏損控(kong)(kong)制(zhi)在北京、上海、鄭州(zhou)等多個(ge)城市得(de)以推廣應用。

3.3 供水用戶服務信息大數據應用

3.3.1 公眾反饋供水問題熱詞與熱圖解析

以公(gong)眾反饋的供水客服、網(wang)絡輿情等數據(ju)作為(wei)數據(ju)應用的核心，通(tong)過解析問(wen)(wen)題熱(re)詞和熱(re)圖(tu)，可精準掌握服務(wu)痛點問(wen)(wen)題與(yu)公(gong)眾輿情。進(jin)一步結合與(yu)之相(xiang)關(guan)的生產(chan)、營銷等數據(ju)，開(kai)展問(wen)(wen)題溯(su)源，可為(wei)改(gai)進(jin)服務(wu)提供決策參考依(yi)據(ju)。

以改(gai)進水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)投(tou)訴問(wen)題為例，濟(ji)南水(shui)(shui)(shui)(shui)(shui)(shui)務集(ji)團有(you)(you)限公(gong)司發(fa)(fa)現2016年(nian)10～12月(yue)間，水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)投(tou)訴類(lei)(lei)(lei)客服工(gong)單(dan)數(shu)量增加明顯。通過篩選統計高(gao)(gao)頻詞匯，并(bing)根據(ju)文本語義與(yu)組織結(jie)構進行最(zui)小串分詞，選取高(gao)(gao)頻排序優先的(de)(de)(de)(de)關(guan)鍵詞構建出熱詞庫，發(fa)(fa)現高(gao)(gao)頻熱詞為龍頭水(shui)(shui)(shui)(shui)(shui)(shui)有(you)(you)異(yi)味(wei)(wei)(wei)。將所有(you)(you)涉及(ji)龍頭水(shui)(shui)(shui)(shui)(shui)(shui)有(you)(you)異(yi)味(wei)(wei)(wei)數(shu)據(ju)的(de)(de)(de)(de)發(fa)(fa)生地點(dian)進行數(shu)據(ju)抽取，發(fa)(fa)現所涉及(ji)地點(dian)圍(wei)繞(rao)某道(dao)路周(zhou)邊沿線分布，初步判定投(tou)訴問(wen)題與(yu)該(gai)道(dao)路對(dui)(dui)應供(gong)水(shui)(shui)(shui)(shui)(shui)(shui)廠的(de)(de)(de)(de)出廠水(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)或其原水(shui)(shui)(shui)(shui)(shui)(shui)水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)相(xiang)關(guan)。進一步的(de)(de)(de)(de)相(xiang)關(guan)性(xing)(xing)分析(xi)結(jie)果顯示，原水(shui)(shui)(shui)(shui)(shui)(shui)的(de)(de)(de)(de)藻類(lei)(lei)(lei)物(wu)質(zhi)(zhi)(zhi)濃度(du)與(yu)龍頭水(shui)(shui)(shui)(shui)(shui)(shui)有(you)(you)異(yi)味(wei)(wei)(wei)數(shu)據(ju)條目數(shu)量存在顯著相(xiang)關(guan)性(xing)(xing)（相(xiang)關(guan)系數(shu)為0.969）。溯(su)源調查發(fa)(fa)現，2016年(nian)10～12月(yue)，南水(shui)(shui)(shui)(shui)(shui)(shui)北調原水(shui)(shui)(shui)(shui)(shui)(shui)注入了水(shui)(shui)(shui)(shui)(shui)(shui)源地水(shui)(shui)(shui)(shui)(shui)(shui)庫，導致水(shui)(shui)(shui)(shui)(shui)(shui)庫水(shui)(shui)(shui)(shui)(shui)(shui)體藻類(lei)(lei)(lei)物(wu)質(zhi)(zhi)(zhi)濃度(du)急(ji)劇升(sheng)高(gao)(gao)，并(bing)分解產生了嗅味(wei)(wei)(wei)物(wu)質(zhi)(zhi)(zhi)，造(zao)成水(shui)(shui)(shui)(shui)(shui)(shui)體異(yi)味(wei)(wei)(wei)增加。經此分析(xi)，供(gong)水(shui)(shui)(shui)(shui)(shui)(shui)單(dan)位在外水(shui)(shui)(shui)(shui)(shui)(shui)注入水(shui)(shui)(shui)(shui)(shui)(shui)庫之際(ji)，增加了對(dui)(dui)藻類(lei)(lei)(lei)指標的(de)(de)(de)(de)檢測，提前制(zhi)(zhi)定生產預案控制(zhi)(zhi)水(shui)(shui)(shui)(shui)(shui)(shui)質(zhi)(zhi)(zhi)異(yi)味(wei)(wei)(wei)，減(jian)少(shao)了此類(lei)(lei)(lei)問(wen)題的(de)(de)(de)(de)投(tou)訴率。

3.3.2 供水用戶用水行為分析

以(yi)供(gong)(gong)水客服(fu)數據(ju)中的(de)(de)供(gong)(gong)水用(yong)戶信息數據(ju)與供(gong)(gong)水管網末端(duan)小區二次(ci)供(gong)(gong)水數據(ju)為基礎，結(jie)合其(qi)它相關數據(ju)，采用(yong)適當(dang)的(de)(de)數據(ju)挖掘技術，可構建(jian)二次(ci)供(gong)(gong)水泵(beng)房指標變(bian)量(liang)時間序列，繼而對不同序列之(zhi)間的(de)(de)指標變(bian)量(liang)的(de)(de)趨勢(shi)性(xing)、周期性(xing)及其(qi)隨機性(xing)進行相關相異分析，在此基礎上總結(jie)得出該(gai)序列所代表小區用(yong)水行為變(bian)化趨勢(shi)。

為(wei)優(you)化供(gong)水(shui)(shui)(shui)管網末端(duan)的(de)壓力調度、改善(shan)客戶(hu)(hu)(hu)服務質量，濟南水(shui)(shui)(shui)務集(ji)團有限公司以供(gong)水(shui)(shui)(shui)用戶(hu)(hu)(hu)信息與(yu)2018年(nian)(nian)(nian)的(de)二次供(gong)水(shui)(shui)(shui)生產數(shu)(shu)(shu)據(ju)為(wei)基礎，結(jie)合(he)與(yu)之相關的(de)其它業務數(shu)(shu)(shu)據(ju)，分析了(le)(le)不同(tong)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)之間(jian)(jian)的(de)用水(shui)(shui)(shui)量變化特(te)征(zheng)。以其中2座(zuo)泵(beng)房(fang)為(wei)例，泵(beng)房(fang)a所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)建(jian)成于(yu)(yu)2004年(nian)(nian)(nian)、泵(beng)房(fang)b所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)建(jian)成于(yu)(yu)2011年(nian)(nian)(nian)，兩座(zuo)泵(beng)房(fang)設(she)備運行狀(zhuang)態良(liang)好，在(zai)(zai)(zai)線儀表經過校準，且上(shang)游供(gong)水(shui)(shui)(shui)水(shui)(shui)(shui)廠相同(tong)、上(shang)級加壓站相同(tong)。分析發現(xian)，泵(beng)房(fang)a實(shi)時(shi)(shi)流(liu)量最高值多出(chu)(chu)現(xian)在(zai)(zai)(zai)0時(shi)(shi)，18時(shi)(shi)，7時(shi)(shi)，9時(shi)(shi)，10時(shi)(shi)，15時(shi)(shi)，16時(shi)(shi)，最低值多出(chu)(chu)現(xian)在(zai)(zai)(zai)5時(shi)(shi)，23時(shi)(shi)，2時(shi)(shi)，全天用水(shui)(shui)(shui)無明(ming)顯(xian)高峰(feng)；泵(beng)房(fang)b實(shi)時(shi)(shi)流(liu)量最高值多出(chu)(chu)現(xian)在(zai)(zai)(zai)0，22，7時(shi)(shi)，最低值多出(chu)(chu)現(xian)在(zai)(zai)(zai)6時(shi)(shi)，23時(shi)(shi)，3時(shi)(shi)，用水(shui)(shui)(shui)量的(de)峰(feng)谷波動明(ming)顯(xian)。結(jie)合(he)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)業主用戶(hu)(hu)(hu)平均(jun)年(nian)(nian)(nian)齡(ling)（泵(beng)房(fang)a所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)平均(jun)48.5歲，泵(beng)房(fang)b所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)平均(jun)38.3歲），推斷(duan)出(chu)(chu)泵(beng)房(fang)a所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)，由(you)于(yu)(yu)成年(nian)(nian)(nian)人，尤其老年(nian)(nian)(nian)人較多，用水(shui)(shui)(shui)無明(ming)顯(xian)高峰(feng)；泵(beng)房(fang)b所(suo)在(zai)(zai)(zai)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)，由(you)于(yu)(yu)青壯年(nian)(nian)(nian)、學前及義務教育(yu)階(jie)段(duan)適齡(ling)人群較多，導致出(chu)(chu)現(xian)晨間(jian)(jian)與(yu)晚間(jian)(jian)用水(shui)(shui)(shui)高峰(feng)時(shi)(shi)間(jian)(jian)的(de)波動。根據(ju)不同(tong)小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)之間(jian)(jian)供(gong)水(shui)(shui)(shui)用戶(hu)(hu)(hu)平均(jun)年(nian)(nian)(nian)齡(ling)、小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)位置、小(xiao)(xiao)區(qu)(qu)(qu)(qu)(qu)建(jian)造時(shi)(shi)間(jian)(jian)等因素，針對(dui)性地定性溯源用水(shui)(shui)(shui)行為(wei)特(te)征(zheng)產生原因，從而(er)為(wei)提前做好客戶(hu)(hu)(hu)服務預案和(he)供(gong)水(shui)(shui)(shui)調度決策提供(gong)了(le)(le)數(shu)(shu)(shu)據(ju)支撐。

針對當前城(cheng)市供水監管(guan)中存在的(de)信息化水平較低、數(shu)(shu)(shu)據(ju)(ju)價(jia)(jia)值挖(wa)掘(jue)不(bu)足等問題，基(ji)于(yu)大數(shu)(shu)(shu)據(ju)(ju)分(fen)析(xi)中的(de)數(shu)(shu)(shu)據(ju)(ju)挖(wa)掘(jue)和綜(zong)合評(ping)價(jia)(jia)等技術(shu)，利用(yong)相關(guan)性分(fen)析(xi)、隨機森(sen)林(lin)、神經(jing)網絡(luo)等數(shu)(shu)(shu)據(ju)(ju)分(fen)析(xi)算法，可對城(cheng)市供水大數(shu)(shu)(shu)據(ju)(ju)潛在信息進行(xing)提取分(fen)析(xi)，開展大數(shu)(shu)(shu)據(ju)(ju)技術(shu)在供水安(an)全動態監管(guan)與風(feng)險預警領域的(de)應用(yong)。