國家(jia)發展(zhan)改革委(wei)國家(jia)能源局關于推進“人工智能+”能源高(gao)質量發展(zhan)的實施意見

國能發科(ke)技〔2025〕73號

各(ge)省(sheng)、自(zi)治區、直轄市(shi)及(ji)計(ji)劃(hua)單列市(shi)、新疆生產建設兵(bing)團發展改革委、能源局，有關(guan)中央(yang)企業，有關(guan)行業協會：

為(wei)深(shen)入貫徹黨(dang)中(zhong)央(yang)、國(guo)務院關于發(fa)展(zhan)人工(gong)智(zhi)能的決策部署，落實《國(guo)務院關于深(shen)入實施“人工(gong)智(zhi)能+”行(xing)動的意(yi)見(jian)》（國(guo)發(fa)〔2025〕11號）有關工(gong)作要(yao)求，搶抓人工(gong)智(zhi)能發(fa)展(zhan)重大戰略機遇，突(tu)出(chu)應用(yong)導向(xiang)，加快推動人工(gong)智(zhi)能與(yu)能源(yuan)產(chan)業深(shen)度融合，支撐能源(yuan)高質量發(fa)展(zhan)和高水平(ping)安全(quan)，現提出(chu)如下意(yi)見(jian)。

一、總體要求

堅持以習(xi)近平(ping)新(xin)時代中國特色(se)(se)社會主義(yi)思想為(wei)指導，深(shen)入貫(guan)徹黨的二(er)十(shi)大和二(er)十(shi)屆二(er)中、三中全會精神(shen)，全面貫(guan)徹習(xi)近平(ping)總書(shu)記關于推動人(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)與實體(ti)經濟深(shen)度融(rong)合(he)(he)、培育(yu)壯大智(zhi)(zhi)(zhi)能(neng)(neng)(neng)產(chan)業的重要(yao)指示精神(shen)，以拓展(zhan)人(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)與能(neng)(neng)(neng)源(yuan)(yuan)領域深(shen)度融(rong)合(he)(he)應(ying)(ying)用(yong)(yong)場(chang)景為(wei)重要(yao)依托，以提(ti)升(sheng)能(neng)(neng)(neng)源(yuan)(yuan)領域人(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)創(chuang)新(xin)應(ying)(ying)用(yong)(yong)技術水平(ping)為(wei)主攻方向，以推進(jin)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)算力(li)(li)與電力(li)(li)協(xie)同發展(zhan)為(wei)必要(yao)支撐，以健全能(neng)(neng)(neng)源(yuan)(yuan)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)化發展(zhan)的創(chuang)新(xin)體(ti)系(xi)為(wei)關鍵保(bao)障(zhang)，著力(li)(li)提(ti)升(sheng)能(neng)(neng)(neng)源(yuan)(yuan)系(xi)統安全可靠與靈活高效運行能(neng)(neng)(neng)力(li)(li)，保(bao)障(zhang)能(neng)(neng)(neng)源(yuan)(yuan)安全穩(wen)定供(gong)應(ying)(ying)和綠色(se)(se)低碳轉型(xing)，加快培育(yu)新(xin)質生產(chan)力(li)(li)，為(wei)新(xin)型(xing)能(neng)(neng)(neng)源(yuan)(yuan)體(ti)系(xi)建設提(ti)供(gong)有力(li)(li)支撐。

到(dao)2027年(nian)，能源(yuan)與(yu)人工(gong)(gong)智(zhi)(zhi)能融合(he)創新體(ti)系(xi)(xi)初(chu)步構建，算力(li)與(yu)電力(li)協(xie)同發(fa)(fa)展根(gen)基不斷夯實，人工(gong)(gong)智(zhi)(zhi)能賦(fu)能能源(yuan)核心(xin)技(ji)術(shu)取得顯著突破，應(ying)(ying)(ying)用更加廣(guang)泛(fan)深(shen)入。推(tui)動五個(ge)(ge)以(yi)上專業大模(mo)型(xing)(xing)在電網、發(fa)(fa)電、煤炭、油氣等行業深(shen)度應(ying)(ying)(ying)用，挖掘十(shi)個(ge)(ge)以(yi)上可復制、易推(tui)廣(guang)、有競爭力(li)的重點示(shi)范(fan)項目，探索(suo)(suo)百(bai)個(ge)(ge)典型(xing)(xing)應(ying)(ying)(ying)用場景賦(fu)能路徑，培育一批能源(yuan)行業人工(gong)(gong)智(zhi)(zhi)能技(ji)術(shu)應(ying)(ying)(ying)用研(yan)發(fa)(fa)創新平臺，制定完善百(bai)項技(ji)術(shu)標準，培養一批能源(yuan)與(yu)人工(gong)(gong)智(zhi)(zhi)能復合(he)型(xing)(xing)人才(cai)，探索(suo)(suo)建立(li)能源(yuan)領域(yu)人工(gong)(gong)智(zhi)(zhi)能技(ji)術(shu)研(yan)發(fa)(fa)應(ying)(ying)(ying)用金融支撐體(ti)系(xi)(xi)，形(xing)成符合(he)我國(guo)國(guo)情的能源(yuan)領域(yu)人工(gong)(gong)智(zhi)(zhi)能技(ji)術(shu)創新發(fa)(fa)展模(mo)式，能源(yuan)領域(yu)智(zhi)(zhi)能化成效初(chu)顯。

到2030年，能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)領(ling)域人工(gong)(gong)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)專(zhuan)用技術(shu)與(yu)應(ying)用總體達到世界領(ling)先水(shui)平。算力電力協同(tong)機制進一步(bu)完善，建立綠色、經濟、安全、高效的(de)算力用能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)模(mo)式。能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)與(yu)人工(gong)(gong)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)融合(he)的(de)理(li)論(lun)與(yu)技術(shu)創(chuang)新取得(de)(de)明顯成(cheng)(cheng)效，能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)領(ling)域人工(gong)(gong)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)技術(shu)實(shi)(shi)現(xian)跨(kua)領(ling)域、跨(kua)行業、跨(kua)業務場景賦能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)，在(zai)電力智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)調控、能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)資源(yuan)(yuan)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)勘探、新能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)預測等方向(xiang)取得(de)(de)突破，具身智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)、科學智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)等在(zai)關(guan)鍵(jian)場景實(shi)(shi)現(xian)落地(di)應(ying)用。形成(cheng)(cheng)一批全球領(ling)先的(de)“人工(gong)(gong)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)+”能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)相關(guan)研發創(chuang)新平臺和復(fu)合(he)人才培養基地(di)，建成(cheng)(cheng)更加完善的(de)政(zheng)策(ce)體系，持續引導“人工(gong)(gong)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)+”能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)高效、健(jian)康、有序創(chuang)新，為能(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)高質(zhi)量發展奠(dian)定堅(jian)實(shi)(shi)基礎。

二、加快能源應用(yong)場(chang)景賦能

（一）人(ren)(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)+電(dian)(dian)(dian)(dian)網。圍繞新(xin)型(xing)(xing)電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)系統下的電(dian)(dian)(dian)(dian)網安全(quan)、新(xin)能(neng)(neng)(neng)源消納(na)、運行(xing)效率等要(yao)求，開展電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)供需預測(ce)(ce)、電(dian)(dian)(dian)(dian)網智(zhi)(zhi)(zhi)能(neng)(neng)(neng)診(zhen)斷分(fen)析、規劃(hua)方案智(zhi)(zhi)(zhi)能(neng)(neng)(neng)生成等電(dian)(dian)(dian)(dian)網規劃(hua)設計應(ying)(ying)用，加強電(dian)(dian)(dian)(dian)網工程(cheng)智(zhi)(zhi)(zhi)慧建(jian)設管理；推(tui)進電(dian)(dian)(dian)(dian)網多尺(chi)度(du)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)仿真分(fen)析，探索人(ren)(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)模型(xing)(xing)在電(dian)(dian)(dian)(dian)網智(zhi)(zhi)(zhi)能(neng)(neng)(neng)輔(fu)助決策(ce)和(he)(he)調度(du)控制(zhi)方面的應(ying)(ying)用，提(ti)(ti)升電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)系統源網荷儲全(quan)要(yao)素(su)安全(quan)可靠低(di)碳運行(xing)水(shui)(shui)平(ping)；穩步(bu)提(ti)(ti)高輸變電(dian)(dian)(dian)(dian)等關鍵裝備(bei)(bei)研制(zhi)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)化水(shui)(shui)平(ping)；推(tui)動電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)設備(bei)(bei)故障預測(ce)(ce)性(xing)維護，打(da)造具備(bei)(bei)自(zi)主感知、決策(ce)、執行(xing)能(neng)(neng)(neng)力(li)(li)(li)(li)的電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)設備(bei)(bei)健康(kang)管理智(zhi)(zhi)(zhi)能(neng)(neng)(neng)體，提(ti)(ti)升設備(bei)(bei)精(jing)益化管理水(shui)(shui)平(ping)；推(tui)動營配調智(zhi)(zhi)(zhi)能(neng)(neng)(neng)一體化應(ying)(ying)用，構建(jian)電(dian)(dian)(dian)(dian)網運營服(fu)務智(zhi)(zhi)(zhi)能(neng)(neng)(neng)支撐體系，提(ti)(ti)升電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)客戶全(quan)過程(cheng)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)服(fu)務水(shui)(shui)平(ping)；促進人(ren)(ren)工智(zhi)(zhi)(zhi)能(neng)(neng)(neng)技術融入電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)應(ying)(ying)急體系和(he)(he)能(neng)(neng)(neng)力(li)(li)(li)(li)建(jian)設，提(ti)(ti)升電(dian)(dian)(dian)(dian)力(li)(li)(li)(li)系統防災(zai)(zai)減災(zai)(zai)救災(zai)(zai)智(zhi)(zhi)(zhi)能(neng)(neng)(neng)化水(shui)(shui)平(ping)。

專欄1 人工智能+電網典型應用場景

電(dian)網智(zhi)能(neng)(neng)(neng)(neng)規劃(hua)設(she)計(ji)與生產建(jian)(jian)設(she)。構建(jian)(jian)電(dian)力供需智(zhi)能(neng)(neng)(neng)(neng)預測(ce)、電(dian)網運(yun)行智(zhi)能(neng)(neng)(neng)(neng)診斷(duan)分析、電(dian)網規劃(hua)智(zhi)能(neng)(neng)(neng)(neng)輔助決策、輸變(bian)電(dian)設(she)施智(zhi)能(neng)(neng)(neng)(neng)設(she)計(ji)等應(ying)(ying)用，應(ying)(ying)用人工(gong)智(zhi)能(neng)(neng)(neng)(neng)技(ji)(ji)術(shu)開(kai)展規劃(hua)設(she)計(ji)和(he)技(ji)(ji)術(shu)經(jing)濟(ji)分析，推動電(dian)網規劃(hua)設(she)計(ji)作(zuo)業(ye)模(mo)式向(xiang)智(zhi)能(neng)(neng)(neng)(neng)化(hua)轉(zhuan)變(bian)。聚焦建(jian)(jian)設(she)階段(duan)的作(zuo)業(ye)感知與業(ye)務監(jian)(jian)測(ce)，構建(jian)(jian)電(dian)網建(jian)(jian)設(she)的人工(gong)智(zhi)能(neng)(neng)(neng)(neng)違章識別、進度(du)仿真、在線(xian)監(jian)(jian)測(ce)、管控指標實(shi)時分析、作(zuo)業(ye)流程智(zhi)能(neng)(neng)(neng)(neng)管理等應(ying)(ying)用，促(cu)進電(dian)網工(gong)程建(jian)(jian)造(zao)智(zhi)能(neng)(neng)(neng)(neng)升(sheng)級。

電(dian)網調度(du)運行。在(zai)全國統一電(dian)力(li)市場(chang)建(jian)設背景(jing)下，構建(jian)新能源功率預測、負荷預測、離(li)線仿真分析、在(zai)線安全分析、極端應(ying)急處置、調度(du)輔助決(jue)策、市場(chang)出清(qing)運籌(chou)優化(hua)、電(dian)力(li)市場(chang)智慧(hui)決(jue)策等(deng)方面的(de)智能化(hua)應(ying)用，持續完(wan)善新一代智能調控(kong)技術(shu)支(zhi)持體系，支(zhi)撐新型電(dian)力(li)系統安全穩(wen)定(ding)運行。

電力設(she)(she)(she)備(bei)狀態評價與智能(neng)運維(wei)。構建設(she)(she)(she)備(bei)狀態智能(neng)感知與預警、設(she)(she)(she)備(bei)故障智能(neng)定(ding)位與診斷、設(she)(she)(she)備(bei)狀態檢修智能(neng)決策、設(she)(she)(she)備(bei)災害風險智能(neng)預測、檢修工作票智能(neng)生成等應用，提升設(she)(she)(she)備(bei)精益化管理水平。

配(pei)電(dian)(dian)網(wang)智能(neng)運行管理。構(gou)建配(pei)電(dian)(dian)網(wang)實時感知、風險分析、智能(neng)決(jue)策(ce)等技術應用，全面(mian)提升配(pei)電(dian)(dian)網(wang)智慧控制能(neng)力和供電(dian)(dian)可靠性，加強(qiang)配(pei)電(dian)(dian)網(wang)層面(mian)源網(wang)荷儲協(xie)同調控。

電(dian)力應(ying)急搶修(xiu)。構(gou)建(jian)電(dian)力系(xi)統災害(hai)風險智能(neng)預警(jing)、損毀(hui)情況智能(neng)分析、應(ying)急方案(an)智能(neng)決(jue)策等輔助(zhu)決(jue)策系(xi)統，推進電(dian)力應(ying)急搶修(xiu)技術裝備智能(neng)化(hua)應(ying)用，提升電(dian)力系(xi)統防災減(jian)災救災能(neng)力。

（二(er)）人(ren)(ren)工(gong)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)+能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)新(xin)(xin)業態。圍繞(rao)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)保供(gong)和綠色低(di)碳轉型(xing)需求，推進人(ren)(ren)工(gong)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)技(ji)術在(zai)(zai)虛擬電(dian)廠（含負(fu)(fu)荷聚合(he)商）、分布式儲能(neng)(neng)(neng)(neng)(neng)(neng)(neng)、電(dian)動(dong)(dong)(dong)汽車車網互動(dong)(dong)(dong)等靈活(huo)性調(diao)節資源(yuan)中(zhong)的應用，提(ti)升(sheng)負(fu)(fu)荷側(ce)群控(kong)優(you)化(hua)和動(dong)(dong)(dong)態響應能(neng)(neng)(neng)(neng)(neng)(neng)(neng)力(li)；加強人(ren)(ren)工(gong)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)技(ji)術在(zai)(zai)新(xin)(xin)型(xing)儲能(neng)(neng)(neng)(neng)(neng)(neng)(neng)與(yu)電(dian)力(li)系(xi)統(tong)協同優(you)化(hua)調(diao)度(du)以(yi)及全(quan)(quan)生(sheng)命周期安全(quan)(quan)中(zhong)的應用，推動(dong)(dong)(dong)可再(zai)生(sheng)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)制氫(qing)生(sheng)產工(gong)藝智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)尋優(you)。強化(hua)人(ren)(ren)工(gong)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)技(ji)術賦能(neng)(neng)(neng)(neng)(neng)(neng)(neng)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)生(sheng)產過程中(zhong)的節能(neng)(neng)(neng)(neng)(neng)(neng)(neng)和碳排放(fang)管(guan)理，提(ti)升(sheng)多(duo)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)互補綜合(he)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)系(xi)統(tong)電(dian)、熱、冷、氣聯(lian)供(gong)的綜合(he)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)效和降碳水平(ping)。推動(dong)(dong)(dong)人(ren)(ren)工(gong)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)在(zai)(zai)零碳園區、智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)微電(dian)網、算(suan)電(dian)協同中(zhong)的應用，提(ti)升(sheng)源(yuan)網荷儲一體(ti)化(hua)智能(neng)(neng)(neng)(neng)(neng)(neng)(neng)運行水平(ping)，促進新(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)就地消(xiao)納。

專欄2 人工智能+能源新業態典型應用場景

虛擬電(dian)廠精準控制(zhi)與(yu)智能運營。虛擬電(dian)廠運營商平臺根據(ju)電(dian)網調節(jie)指令、市(shi)場(chang)信息，結(jie)合(he)資源特性的(de)動態變(bian)化，進行控制(zhi)策略的(de)智能優化和控制(zhi)指令的(de)智能生成，實(shi)現大規模靈(ling)活(huo)性資源聚合(he)優化調控、實(shi)現虛擬電(dian)廠參與(yu)電(dian)力市(shi)場(chang)的(de)智慧(hui)交易(yi)決策。

綠氫生產工(gong)藝智能(neng)尋優。融合(he)風(feng)光功率波動預測、儲氫罐容量、電解(jie)槽(cao)溫度、催化(hua)劑狀態等(deng)多維數據，基于人工(gong)智能(neng)算法，智能(neng)驅動電解(jie)槽(cao)電流密度動態尋優，構建(jian)電解(jie)制氫-儲氫-用氫全鏈條智能(neng)調控系統，實現可再生能(neng)源功率波動與(yu)電解(jie)裝置柔性負荷的(de)毫(hao)秒級匹配。

園區(qu)(qu)智能(neng)降碳(tan)。基于光伏、儲(chu)能(neng)等(deng)設備(bei)運行數據，園區(qu)(qu)智能(neng)降碳(tan)協同控制系統實(shi)時動態優化(hua)能(neng)源調(diao)度策略(lve)，結合電價與碳(tan)排放因子(zi)自動調(diao)節空調(diao)溫度、充電樁功率(lv)及設備(bei)啟停時序，通過(guo)增強現實(shi)可(ke)視化(hua)界面和語音助手向(xiang)用(yong)戶推(tui)送個性化(hua)節能(neng)建議，形成“碳(tan)-能(neng)-費”智能(neng)協同模(mo)式。

新(xin)型儲能(neng)(neng)(neng)智(zhi)能(neng)(neng)(neng)化(hua)運(yun)行(xing)。針對(dui)新(xin)型儲能(neng)(neng)(neng)動態適配電力(li)(li)系統(tong)調度、廣域協(xie)同互動、弱電網支(zhi)撐、電池裝(zhuang)備安全(quan)(quan)監(jian)測、設備本體(ti)評估與(yu)運(yun)維，通過人工(gong)智(zhi)能(neng)(neng)(neng)技(ji)術，提升面(mian)向弱電網的(de)多類型儲能(neng)(neng)(neng)協(xie)調控制能(neng)(neng)(neng)力(li)(li)，構建(jian)新(xin)能(neng)(neng)(neng)源與(yu)配建(jian)新(xin)型儲能(neng)(neng)(neng)廣域協(xie)同優化(hua)控制、儲能(neng)(neng)(neng)電站(zhan)智(zhi)能(neng)(neng)(neng)評估、智(zhi)慧(hui)運(yun)維決策支(zhi)持、全(quan)(quan)生命周(zhou)期安全(quan)(quan)等應用(yong)體(ti)系，提升系統(tong)友好型新(xin)能(neng)(neng)(neng)源電站(zhan)的(de)電力(li)(li)供(gong)應保(bao)障能(neng)(neng)(neng)力(li)(li)。

智(zhi)能(neng)(neng)(neng)(neng)營銷服(fu)(fu)務(wu)(wu)。針(zhen)對(dui)油、氣、電等直接面向(xiang)客(ke)戶服(fu)(fu)務(wu)(wu)場景，構建座席(xi)業務(wu)(wu)受理智(zhi)能(neng)(neng)(neng)(neng)輔助、智(zhi)能(neng)(neng)(neng)(neng)客(ke)戶服(fu)(fu)務(wu)(wu)、供電方案智(zhi)能(neng)(neng)(neng)(neng)生(sheng)成、綜合用(yong)能(neng)(neng)(neng)(neng)方案智(zhi)能(neng)(neng)(neng)(neng)生(sheng)成、運維(wei)工單(dan)智(zhi)能(neng)(neng)(neng)(neng)派發、用(yong)戶用(yong)能(neng)(neng)(neng)(neng)異常診(zhen)斷等智(zhi)能(neng)(neng)(neng)(neng)化(hua)應用(yong)，打(da)造交互式、伴隨式的客(ke)服(fu)(fu)新模(mo)式，提升(sheng)客(ke)戶全過程智(zhi)能(neng)(neng)(neng)(neng)化(hua)服(fu)(fu)務(wu)(wu)水平。

（三）人工(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)+新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)。針對新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)出力(li)(li)波動(dong)性(xing)與間(jian)歇(xie)性(xing)的問題，加快在(zai)高精度功率(lv)預測(ce)(ce)、電力(li)(li)市場、場站智(zhi)慧(hui)運營、新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)規劃(hua)、項目后評價等方(fang)向的人工(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)應用，持續推(tui)動(dong)新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)關(guan)鍵材料及產品不斷迭代和創(chuang)新(xin)(xin)(xin)，推(tui)動(dong)復雜(za)場景(jing)及轉折性(xing)天氣下(xia)功率(lv)預測(ce)(ce)大(da)模型在(zai)更小尺度、更高精準(zhun)度方(fang)向發(fa)(fa)展(zhan)，支撐廣域(yu)新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)資源(yuan)(yuan)(yuan)(yuan)協同(tong)優(you)化，促進(jin)偏遠地區(qu)新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)場站智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)運維發(fa)(fa)展(zhan)，打造(zao)“氣象預測(ce)(ce)+功率(lv)預測(ce)(ce)+智(zhi)慧(hui)交(jiao)易+智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)運維”一體化新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)生產模式，全力(li)(li)支撐新(xin)(xin)(xin)能(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)(yuan)(yuan)(yuan)穩定(ding)供給。

專欄3 人工智能+新能源典型應用場景

氣象預報(bao)與新(xin)能源功(gong)(gong)率(lv)精準(zhun)預測。構建以多時空尺度氣象預報(bao)為核心(xin)的(de)氣象服務體系(xi)，建立氣象-功(gong)(gong)率(lv)非(fei)線性(xing)關系(xi)精準(zhun)挖掘與解析(xi)的(de)多場景(jing)多周期算法大模型，實現新(xin)能源功(gong)(gong)率(lv)精準(zhun)預測。

偏遠地區場站(zhan)智能(neng)(neng)運(yun)維。利用大(da)模型、聲(sheng)紋檢測(ce)、遙(yao)感、機器人(ren)、智能(neng)(neng)穿戴(dai)設(she)備等(deng)技術裝備，實時監測(ce)周邊環(huan)境(jing)及設(she)備運(yun)行狀態(tai)，實現無(wu)(wu)人(ren)機、無(wu)(wu)人(ren)車、無(wu)(wu)人(ren)船、智能(neng)(neng)控制等(deng)多系統智能(neng)(neng)聯動，提升(sheng)設(she)備巡檢效率，提高場站(zhan)的綜(zong)合運(yun)營效率。

新能源規劃設(she)計(ji)。綜合考慮(lv)發電(dian)效率、投資回報率等因素，構建智能化(hua)推薦引擎，提(ti)供最(zui)優機型(xing)匹配(pei)方(fang)(fang)案(an)。融合大模(mo)型(xing)與設(she)計(ji)軟件(jian)，快速(su)生成多版本設(she)計(ji)方(fang)(fang)案(an)并評(ping)估關鍵(jian)參數，提(ti)升設(she)計(ji)效率與質(zhi)量。

智慧工(gong)(gong)地建設。推動人工(gong)(gong)智能技術深度融入工(gong)(gong)程(cheng)建設方案選擇、人員(yuan)管理(li)、風險預警、工(gong)(gong)期管控等電力建設工(gong)(gong)程(cheng)管理(li)全(quan)流程(cheng)，研發無人機巡檢系統(tong)、風險自動研判預警系統(tong)等，實(shi)時捕捉施工(gong)(gong)人員(yuan)違章行為，構建貫穿施工(gong)(gong)全(quan)過程(cheng)的(de)“智慧工(gong)(gong)地”管理(li)平臺，助力提升電力建設工(gong)(gong)程(cheng)安(an)全(quan)質量總體水平。

（四）人工(gong)(gong)智(zhi)(zhi)能(neng)(neng)+水(shui)(shui)(shui)電(dian)。聚焦高海拔高寒(han)地區水(shui)(shui)(shui)電(dian)工(gong)(gong)程(cheng)(cheng)智(zhi)(zhi)能(neng)(neng)化(hua)(hua)建(jian)設與流域水(shui)(shui)(shui)電(dian)站(zhan)群智(zhi)(zhi)慧(hui)調(diao)度(du)運(yun)(yun)營，推進人工(gong)(gong)智(zhi)(zhi)能(neng)(neng)技術在水(shui)(shui)(shui)電(dian)工(gong)(gong)程(cheng)(cheng)建(jian)設中的應用，提升(sheng)(sheng)水(shui)(shui)(shui)電(dian)工(gong)(gong)程(cheng)(cheng)智(zhi)(zhi)能(neng)(neng)化(hua)(hua)設計施工(gong)(gong)管理(li)水(shui)(shui)(shui)平；推進人工(gong)(gong)智(zhi)(zhi)能(neng)(neng)技術與傳統水(shui)(shui)(shui)文模型、氣(qi)(qi)象模型、大(da)規模水(shui)(shui)(shui)庫調(diao)度(du)技術融(rong)合(he)，提升(sheng)(sheng)氣(qi)(qi)象、水(shui)(shui)(shui)文雙向耦合(he)預(yu)測精度(du)，開展調(diao)度(du)決策(ce)優化(hua)(hua)智(zhi)(zhi)能(neng)(neng)應用建(jian)設；推動知識圖譜、大(da)模型、智(zhi)(zhi)能(neng)(neng)體等技術融(rong)入新(xin)一代水(shui)(shui)(shui)電(dian)智(zhi)(zhi)慧(hui)運(yun)(yun)營大(da)腦(nao)，在水(shui)(shui)(shui)電(dian)站(zhan)智(zhi)(zhi)慧(hui)運(yun)(yun)維與精益檢修、智(zhi)(zhi)能(neng)(neng)大(da)壩態勢感知與智(zhi)(zhi)慧(hui)管理(li)等重點(dian)領域形成智(zhi)(zhi)能(neng)(neng)化(hua)(hua)解決方案。

專欄4 人工智能+水電典型應用場景

智(zhi)能水電(dian)工(gong)(gong)程(cheng)建設(she)。基于多源遙感數據融合(he)和智(zhi)能機器人等(deng)人工(gong)(gong)智(zhi)能技術，建立水電(dian)工(gong)(gong)程(cheng)地質智(zhi)能化(hua)勘測設(she)計體(ti)系(xi)，實現機組設(she)備數字(zi)化(hua)智(zhi)能化(hua)安裝調試(shi)，提(ti)升(sheng)水電(dian)工(gong)(gong)程(cheng)智(zhi)能化(hua)施(shi)工(gong)(gong)管理水平。

氣(qi)象水文(wen)(wen)聯合預(yu)(yu)(yu)測(ce)。基于流(liu)域氣(qi)象水文(wen)(wen)雙向耦(ou)合預(yu)(yu)(yu)測(ce)大(da)模型，構建洪旱(han)極端(duan)事(shi)件風險量化(hua)工具(ju)，充(chong)分融合氣(qi)象知識(shi)、水文(wen)(wen)知識(shi)和流(liu)域地理信息(xi)，提(ti)升氣(qi)象水文(wen)(wen)預(yu)(yu)(yu)報精度和預(yu)(yu)(yu)見期(qi)。

流域(yu)綜合(he)調(diao)度。基于流域(yu)站群聯合(he)智慧(hui)優(you)化(hua)調(diao)度、風險控制(zhi)和(he)模擬仿真等關(guan)鍵技術，建設精準(zhun)調(diao)度決策優(you)化(hua)智能應用，實現(xian)對水資(zi)源調(diao)度方(fang)案執行情(qing)況的實時(shi)監測、分析和(he)評估，在(zai)時(shi)間和(he)空間上對水資(zi)源分配進(jin)行優(you)化(hua)，提(ti)高水能利(li)用率，增加(jia)發電(dian)效(xiao)益。

設備(bei)智能(neng)(neng)(neng)運(yun)檢(jian)。基于物(wu)理(li)(li)場、聲學、視覺、智能(neng)(neng)(neng)傳感器等(deng)(deng)(deng)多源數據(ju)以及知識圖譜(pu)、大模(mo)型等(deng)(deng)(deng)技術，推動水(shui)電關鍵設備(bei)實(shi)現狀態全(quan)息監測、全(quan)生(sheng)命周期健康管理(li)(li)、智能(neng)(neng)(neng)運(yun)維和狀態檢(jian)修等(deng)(deng)(deng)業務領域全(quan)流(liu)程智能(neng)(neng)(neng)化(hua)升級，實(shi)現運(yun)維知識結構化(hua)管理(li)(li)與基于大模(mo)型-智能(neng)(neng)(neng)體的智能(neng)(neng)(neng)輔助決策系統。

大(da)(da)壩高質量運(yun)行(xing)。構建大(da)(da)壩典型(xing)病害特征數據庫與知識圖譜，結合大(da)(da)壩智能(neng)感知-融合-診斷-防控(kong)理論方法，實現多元驅動(dong)的(de)大(da)(da)壩安(an)全(quan)狀態早期識別-自診斷-自適(shi)應預(yu)警-智能(neng)饋(kui)控(kong)，確保水(shui)(shui)電(dian)站大(da)(da)壩運(yun)行(xing)安(an)全(quan)，支撐水(shui)(shui)庫大(da)(da)壩高質量運(yun)行(xing)管理。

（五）人工智(zhi)(zhi)能(neng)(neng)+火電。圍繞火電清潔降碳、安全可靠、高效調(diao)節、智(zhi)(zhi)能(neng)(neng)運(yun)行(xing)的(de)發展(zhan)(zhan)方向，在燃料管(guan)控(kong)、生產運(yun)行(xing)優(you)化與智(zhi)(zhi)能(neng)(neng)控(kong)制(zhi)、設備全生命周(zhou)(zhou)期管(guan)理等(deng)業(ye)務場(chang)景，協同(tong)開(kai)展(zhan)(zhan)人工智(zhi)(zhi)能(neng)(neng)賦能(neng)(neng)及技術創新(xin)。加快火電數字化設計建造和(he)智(zhi)(zhi)能(neng)(neng)化升級，推動火電運(yun)行(xing)控(kong)制(zhi)系統智(zhi)(zhi)能(neng)(neng)化發展(zhan)(zhan)和(he)應用，提升火電關鍵裝備全生命周(zhou)(zhou)期智(zhi)(zhi)能(neng)(neng)監測(ce)及健(jian)康(kang)管(guan)理能(neng)(neng)力(li)，助(zhu)力(li)火電支撐保障(zhang)能(neng)(neng)力(li)進(jin)一步提升。

專欄5 人工智能+火電典型應用場景

燃料(liao)(liao)智(zhi)(zhi)能管控。基于(yu)燃料(liao)(liao)市場價格(ge)波動、庫存(cun)量、耗煤(mei)量以及(ji)煤(mei)堆三維結(jie)構、煤(mei)質分析等(deng)多維度多類型數據，采用先進傳(chuan)感(gan)、圖像識別、規(gui)則理解以及(ji)智(zhi)(zhi)能體等(deng)技術，實(shi)現(xian)燃料(liao)(liao)數量、質量等(deng)智(zhi)(zhi)能檢測和(he)智(zhi)(zhi)能管控。

生(sheng)產運(yun)行優(you)(you)化(hua)(hua)。基于大模(mo)型和生(sheng)產運(yun)營(ying)相(xiang)關系統數據，實現(xian)生(sheng)產運(yun)營(ying)過程(cheng)中燃料摻配、運(yun)行優(you)(you)化(hua)(hua)、智能(neng)靈(ling)活(huo)調(diao)峰、安(an)全(quan)智能(neng)管控(kong)等核心業(ye)務場景智能(neng)化(hua)(hua)升(sheng)級，提升(sheng)生(sheng)產運(yun)營(ying)的智能(neng)化(hua)(hua)水(shui)平和效率。

設(she)備全生命(ming)(ming)周期管理。基于大模型(xing)和(he)機器人(ren)等(deng)人(ren)工智能技術，通過對汽輪機（含燃(ran)氣(qi)輪機）、發電機、鍋(guo)爐受熱面等(deng)關鍵設(she)備多(duo)類型(xing)數據進行實時狀(zhuang)態(tai)監測(ce)(ce)，實現設(she)備狀(zhuang)態(tai)全景監測(ce)(ce)、健(jian)康量化評估、隱(yin)患識別與故障預警、剩余壽(shou)命(ming)(ming)預測(ce)(ce)、運行方案調(diao)整(zheng)、異(yi)常分(fen)析判斷(duan)和(he)隱(yin)患閉環管理。

智能(neng)技術(shu)監(jian)督(du)及評價。依托鍋爐、汽輪機（含燃氣輪機）、發電機等關鍵設備的(de)海(hai)量運行數據與火(huo)電技術(shu)監(jian)督(du)工作(zuo)相關資(zi)料，基于火(huo)電大模(mo)型多(duo)模(mo)態分析能(neng)力，深度融(rong)合火(huo)電特色場景，提升(sheng)技術(shu)監(jian)督(du)的(de)智能(neng)化(hua)和人員專業(ye)能(neng)力。

（六）人(ren)工(gong)智能(neng)+核(he)電(dian)(dian)(dian)。圍繞核(he)電(dian)(dian)(dian)安(an)全發展，構建核(he)電(dian)(dian)(dian)安(an)全預警(jing)、電(dian)(dian)(dian)站運(yun)行事(shi)件智能(neng)溯源分析(xi)、應急響應的(de)智能(neng)輔助支持(chi)系(xi)統(tong)，開展核(he)工(gong)業特種(zhong)運(yun)維機器人(ren)技(ji)術(shu)攻關，持(chi)續推(tui)動(dong)(dong)核(he)電(dian)(dian)(dian)系(xi)統(tong)的(de)自(zi)動(dong)(dong)啟停等(deng)技(ji)術(shu)升級演進(jin)，探索人(ren)工(gong)智能(neng)技(ji)術(shu)助力離子體(ti)預測控制(zhi)、可控核(he)聚變(bian)等(deng)技(ji)術(shu)路徑，推(tui)動(dong)(dong)核(he)電(dian)(dian)(dian)行業向數據驅動(dong)(dong)、模型牽引、智能(neng)管控的(de)新模式穩步(bu)轉型。

專欄6 人工智能+核電典型應用場景

核電智能安全(quan)管控。借(jie)助(zhu)數據(ju)治理及人(ren)工智能技(ji)(ji)術(shu)，聚焦運行事件溯(su)源、技(ji)(ji)術(shu)規格書及運行參數邊(bian)界條件，智能識別人(ren)員、設(she)備、環境的不安全(quan)狀態(tai)，推進(jin)安全(quan)預警、智能應急響(xiang)應等場景技(ji)(ji)術(shu)攻關與(yu)應用。

核(he)電(dian)(dian)智(zhi)能運維(wei)。利(li)用(yong)各(ge)階段的構筑(zhu)物、系(xi)統及(ji)設備/部(bu)件的數據，建立數據驅動(dong)的核(he)電(dian)(dian)廠模(mo)(mo)型，推(tui)動(dong)核(he)電(dian)(dian)人工(gong)智(zhi)能小模(mo)(mo)型及(ji)專業大模(mo)(mo)型研發，推(tui)進人工(gong)智(zhi)能技術(shu)在(zai)核(he)電(dian)(dian)系(xi)統智(zhi)能監(jian)測、預(yu)(yu)警、診(zhen)斷和預(yu)(yu)測中的應用(yong)，提升(sheng)(sheng)機組(zu)性(xing)(xing)能智(zhi)能診(zhen)斷和優化(hua)能力(li)，提升(sheng)(sheng)關鍵(jian)設備、系(xi)統及(ji)機組(zu)的一鍵(jian)啟停等能力(li)，拓展(zhan)高(gao)放射性(xing)(xing)、水下及(ji)密閉空間(jian)等高(gao)危場景機器人作業的范圍與深度。

可控(kong)(kong)核聚變(bian)(bian)(bian)智(zhi)(zhi)能控(kong)(kong)制(zhi)。結合可控(kong)(kong)核聚變(bian)(bian)(bian)裝置多物(wu)理場耦合特征，基于人工智(zhi)(zhi)能技術(shu)開展(zhan)可控(kong)(kong)核聚變(bian)(bian)(bian)智(zhi)(zhi)能控(kong)(kong)制(zhi)系(xi)統研(yan)究，研(yan)發等離(li)子體(ti)位形實時預測(ce)-磁約束參數自適(shi)應調控(kong)(kong)智(zhi)(zhi)能模型，實現托卡馬(ma)克等離(li)子體(ti)穩態運行的(de)智(zhi)(zhi)能化控(kong)(kong)制(zhi)。

（七）人(ren)(ren)工智(zhi)能+煤(mei)(mei)(mei)炭(tan)。聚焦地質勘探、煤(mei)(mei)(mei)礦(kuang)采掘（剝）、煤(mei)(mei)(mei)炭(tan)洗選(xuan)、生(sheng)產(chan)調度(du)、安全管控、設備管理(li)等典型場景，穩(wen)(wen)定獲取復雜地質、多(duo)工況以(yi)及多(duo)時空協(xie)同條件下的(de)各種工況數據，融合應(ying)用智(zhi)能模(mo)型，實(shi)現(xian)生(sheng)產(chan)過程智(zhi)能控制與自主決策，助力(li)少人(ren)(ren)無人(ren)(ren)化作業常態化運行，穩(wen)(wen)步推進(jin)減(jian)人(ren)(ren)、增(zeng)安、提效，進(jin)一步夯實(shi)煤(mei)(mei)(mei)炭(tan)在能源安全中的(de)兜底保障作用。

專欄7 人工智能+煤炭典型應用場景

煤(mei)礦(kuang)地(di)質(zhi)(zhi)勘探(tan)數智(zhi)賦能。基于煤(mei)礦(kuang)專業(ye)大模型，融合地(di)面高精度勘探(tan)與井下動(dong)態智(zhi)能探(tan)測的(de)新技術，構(gou)建(jian)復雜(za)地(di)質(zhi)(zhi)條件下的(de)煤(mei)礦(kuang)地(di)質(zhi)(zhi)數據庫，實現礦(kuang)井地(di)質(zhi)(zhi)信息的(de)全過程動(dong)態協同管理和(he)預警，保障礦(kuang)井高效、快速、綠(lv)色、智(zhi)能生產。

井工(gong)煤(mei)礦采(cai)(cai)(cai)掘(jue)工(gong)藝優化與智(zhi)能控制。通過(guo)多模態(tai)感知(zhi)、大小模型融(rong)合、設(she)(she)備(bei)群(qun)協同(tong)(tong)控制和工(gong)藝動態(tai)優化，挖掘(jue)煤(mei)巖特(te)征信息(xi)，驅動采(cai)(cai)(cai)煤(mei)與掘(jue)進(jin)工(gong)作(zuo)(zuo)面設(she)(she)備(bei)群(qun)智(zhi)能截割、自(zi)主(zhu)(zhu)決策與協同(tong)(tong)控制，實現(xian)采(cai)(cai)(cai)煤(mei)工(gong)作(zuo)(zuo)面生產系(xi)統自(zi)主(zhu)(zhu)運行、掘(jue)進(jin)工(gong)作(zuo)(zuo)面探-掘(jue)-支-錨-運高效協同(tong)(tong)以及少人(ren)無人(ren)化常(chang)態(tai)化作(zuo)(zuo)業，大幅提升采(cai)(cai)(cai)掘(jue)效率和安(an)全水平。

露天(tian)(tian)煤(mei)礦(kuang)(kuang)自(zi)主采(cai)裝與(yu)(yu)運輸無人(ren)化(hua)(hua)。推進(jin)大模(mo)型模(mo)擬爆(bao)破(po)參(can)數與(yu)(yu)穿爆(bao)作(zuo)業(ye)(ye)的融(rong)合，應(ying)用人(ren)工(gong)智能(neng)技術(shu)快速解析采(cai)剝進(jin)度，實現采(cai)-運-排生產(chan)系統內挖掘(jue)機、排土推土機以及(ji)其他輔(fu)助作(zuo)業(ye)(ye)設(she)備常態化(hua)(hua)遠(yuan)控或自(zi)主作(zuo)業(ye)(ye)，以及(ji)礦(kuang)(kuang)用卡車無人(ren)駕駛規模(mo)化(hua)(hua)運行，提升穿爆(bao)智能(neng)化(hua)(hua)程(cheng)度和(he)精(jing)準度，大幅(fu)減少坑下(xia)作(zuo)業(ye)(ye)人(ren)員(yuan)數量，提升露天(tian)(tian)煤(mei)礦(kuang)(kuang)生產(chan)效(xiao)率與(yu)(yu)安(an)全水平(ping)。

煤(mei)炭質(zhi)量(liang)快(kuai)速檢(jian)測(ce)(ce)與(yu)智能洗(xi)選。采集與(yu)構建(jian)(jian)煤(mei)質(zhi)特(te)征數(shu)據(ju)(ju)庫，實(shi)(shi)(shi)時動(dong)態(tai)預測(ce)(ce)煤(mei)炭灰分、硫分、揮發分、水分及(ji)元素含量(liang)等關鍵指標，實(shi)(shi)(shi)現煤(mei)質(zhi)特(te)征智能識(shi)別，大幅提高煤(mei)質(zhi)在線檢(jian)測(ce)(ce)精度，實(shi)(shi)(shi)時反饋(kui)煤(mei)質(zhi)在線檢(jian)測(ce)(ce)數(shu)據(ju)(ju)，優化調(diao)節選煤(mei)生產工藝參數(shu)，提高煤(mei)炭產品質(zhi)量(liang)合格(ge)率(lv)和穩定率(lv)。開(kai)發煤(mei)炭洗(xi)選專(zhuan)業模(mo)型，建(jian)(jian)立工業數(shu)字孿生體，實(shi)(shi)(shi)現煤(mei)炭洗(xi)選全過程的信息動(dong)態(tai)監測(ce)(ce)、趨勢預測(ce)(ce)及(ji)協(xie)同管理。

煤礦重(zhong)大設(she)備狀(zhuang)態(tai)監測和(he)(he)智能(neng)運(yun)(yun)維。建立重(zhong)大設(she)備實時運(yun)(yun)行狀(zhuang)態(tai)和(he)(he)潤(run)滑(hua)、溫震等(deng)檢測數據融合大模(mo)型，實現故障(zhang)診斷和(he)(he)智能(neng)預警，推動煤礦設(she)備預防性檢修，大幅降低故障(zhang)影響生產(chan)時間，有效(xiao)降低維護成本(ben)。

（八）人工智(zhi)(zhi)能(neng)+油(you)氣(qi)。聚焦跨專業協同研(yan)(yan)究、現場(chang)作業操控、生產(chan)運(yun)行(xing)(xing)管(guan)控等(deng)方向，推(tui)動(dong)(dong)勘探地質目標(biao)智(zhi)(zhi)能(neng)評價、開發(fa)方案智(zhi)(zhi)能(neng)優化(hua)、鉆(zhan)井壓裂等(deng)作業參數智(zhi)(zhi)能(neng)調整、煉化(hua)裝(zhuang)置智(zhi)(zhi)能(neng)運(yun)行(xing)(xing)、管(guan)網(wang)運(yun)行(xing)(xing)實(shi)時仿真，加快智(zhi)(zhi)能(neng)鉆(zhan)機、機器人、無人機、智(zhi)(zhi)能(neng)感(gan)知系統(tong)等(deng)智(zhi)(zhi)能(neng)生產(chan)技術裝(zhuang)備的研(yan)(yan)發(fa)與(yu)應用，推(tui)動(dong)(dong)生產(chan)現場(chang)等(deng)全(quan)過程智(zhi)(zhi)能(neng)聯動(dong)(dong)與(yu)自動(dong)(dong)優化(hua)，推(tui)動(dong)(dong)油(you)氣(qi)產(chan)業鏈智(zhi)(zhi)能(neng)化(hua)升級(ji)建設。

專欄8 人工智能+油氣典型應用場景

油氣勘探(tan)智(zhi)(zhi)能(neng)賦能(neng)。提升面(mian)向地震(zhen)、測井、巖心露頭等勘探(tan)專業領(ling)域的(de)軟件智(zhi)(zhi)能(neng)化水平(ping)，構(gou)建面(mian)向地震(zhen)測井處理解釋(shi)的(de)專業大模型，打造面(mian)向有利(li)地質(zhi)目標綜合評價(jia)的(de)智(zhi)(zhi)能(neng)應用系(xi)統，實(shi)現(xian)可控震(zhen)源智(zhi)(zhi)能(neng)輔助駕駛、地震(zhen)檢(jian)波(bo)器埋置(zhi)等機器人示范應用。

油氣藏開發(fa)與(yu)生(sheng)產(chan)智(zhi)(zhi)能(neng)(neng)管(guan)控(kong)。研(yan)(yan)發(fa)油氣開發(fa)數據與(yu)知識智(zhi)(zhi)能(neng)(neng)化(hua)技(ji)術、智(zhi)(zhi)能(neng)(neng)開發(fa)優化(hua)軟(ruan)件和專業大模(mo)型，打造大模(mo)型驅動的(de)協同研(yan)(yan)究與(yu)生(sheng)產(chan)管(guan)理決(jue)策平臺，構建面(mian)向智(zhi)(zhi)慧油氣田開發(fa)生(sheng)產(chan)管(guan)控(kong)的(de)新模(mo)式。

海(hai)(hai)洋油(you)氣生(sheng)產環境預(yu)測(ce)(ce)維護(hu)。聚(ju)焦海(hai)(hai)洋油(you)氣生(sheng)產過程環境保護(hu)和(he)(he)重大(da)風(feng)險防(fang)范、治理等需求，通過生(sheng)產環境智能(neng)監(jian)測(ce)(ce)與異常(chang)預(yu)警、固廢處理智能(neng)管控、溢油(you)智能(neng)識別與應急預(yu)測(ce)(ce)等手段(duan)，形成覆蓋油(you)氣田全域生(sheng)態(tai)(tai)環境狀況的風(feng)險預(yu)知(zhi)、態(tai)(tai)勢感知(zhi)、事(shi)故(gu)早知(zhi)和(he)(he)認知(zhi)決策一體(ti)化能(neng)力。

工(gong)(gong)程技術智能(neng)優化。推(tui)進地面(mian)工(gong)(gong)程智能(neng)設計、鉆井(jing)參(can)數智能(neng)優化、錄井(jing)實時智能(neng)判層、儲層改造及智能(neng)故障診(zhen)斷與風(feng)險評(ping)估，實現井(jing)控機器人示范(fan)應用，保(bao)障復雜地質環境下(xia)施工(gong)(gong)安(an)全高效。

管網仿(fang)真(zhen)及智能調控。推進市(shi)場洞察預(yu)測(ce)(ce)、管網實時(shi)仿(fang)真(zhen)及動態優化、高效智能站庫運行、空天地一(yi)體線路管理(li)及關(guan)鍵(jian)設備監測(ce)(ce)預(yu)警，實現“黑屏”智能調控，提升油氣管網安全生產、油氣保供與公平服務能力。

煉廠生(sheng)(sheng)產(chan)營運(yun)一體化(hua)(hua)優(you)化(hua)(hua)。面向全(quan)流(liu)程計劃優(you)化(hua)(hua)、安全(quan)生(sheng)(sheng)產(chan)智(zhi)能(neng)(neng)識別、設備預(yu)防性維修等環(huan)節，攻(gong)關新(xin)材料研發科(ke)學計算大模型，通過大小模型協同、混合建模等技術(shu)手(shou)段，減(jian)少(shao)工藝波動，降低(di)安全(quan)事故發生(sheng)(sheng)概率(lv)，提升(sheng)生(sheng)(sheng)產(chan)運(yun)營智(zhi)能(neng)(neng)化(hua)(hua)水平。

三、加大關鍵(jian)技術供給

聚(ju)焦(jiao)能源領域數據孤島(dao)化、算(suan)(suan)力(li)碎片化、算(suan)(suan)法黑盒化、算(suan)(suan)力(li)高耗能等技術瓶頸，推動(dong)開展適用(yong)能源領域的數據、算(suan)(suan)力(li)、算(suan)(suan)法等共性關鍵技術攻關。

（一）夯實數(shu)(shu)據(ju)(ju)基礎。針對能(neng)(neng)(neng)源(yuan)領域(yu)高質(zhi)量(liang)數(shu)(shu)據(ju)(ju)集構建和數(shu)(shu)據(ju)(ju)安全需求，推動數(shu)(shu)據(ju)(ju)智能(neng)(neng)(neng)標注、智能(neng)(neng)(neng)增強、數(shu)(shu)據(ju)(ju)合成等技(ji)術(shu)(shu)(shu)應用，推進能(neng)(neng)(neng)源(yuan)數(shu)(shu)據(ju)(ju)分類分級技(ji)術(shu)(shu)(shu)、隱(yin)私計算技(ji)術(shu)(shu)(shu)以及智能(neng)(neng)(neng)數(shu)(shu)據(ju)(ju)動態加(jia)密和跨域(yu)可(ke)信溯源(yuan)等技(ji)術(shu)(shu)(shu)研發，優化數(shu)(shu)據(ju)(ju)分享機制，加(jia)快形成能(neng)(neng)(neng)源(yuan)領域(yu)高質(zhi)量(liang)數(shu)(shu)據(ju)(ju)集，確(que)保(bao)能(neng)(neng)(neng)源(yuan)數(shu)(shu)據(ju)(ju)全流程(cheng)安全可(ke)靠。

（二）強化算(suan)(suan)力(li)(li)(li)(li)(li)支撐。針對能源領域租建結(jie)合模式下的多(duo)元(yuan)異(yi)(yi)構(gou)算(suan)(suan)力(li)(li)(li)(li)(li)融(rong)合利(li)用需(xu)求，開展(zhan)多(duo)元(yuan)異(yi)(yi)構(gou)算(suan)(suan)力(li)(li)(li)(li)(li)統一調(diao)度、任務智(zhi)能編排(pai)、存算(suan)(suan)網(wang)一體(ti)化融(rong)合、算(suan)(suan)力(li)(li)(li)(li)(li)池化等關鍵技術攻關，提升智(zhi)算(suan)(suan)服務水平。持續開展(zhan)能源算(suan)(suan)力(li)(li)(li)(li)(li)需(xu)求監測，統籌規劃算(suan)(suan)力(li)(li)(li)(li)(li)、電力(li)(li)(li)(li)(li)和通(tong)信網(wang)絡資源，構(gou)建算(suan)(suan)力(li)(li)(li)(li)(li)、電力(li)(li)(li)(li)(li)深(shen)度融(rong)合的算(suan)(suan)電協同發展(zhan)機制，不(bu)斷提高算(suan)(suan)力(li)(li)(li)(li)(li)中心(xin)綠電比(bi)例。

（三）提升(sheng)模型(xing)基礎能(neng)力(li)(li)。針對能(neng)源(yuan)(yuan)領域(yu)對于(yu)模型(xing)安(an)(an)全(quan)性和可(ke)解釋(shi)(shi)性的(de)需求(qiu)，推動模型(xing)算(suan)法(fa)、應(ying)(ying)用(yong)系統等(deng)安(an)(an)全(quan)能(neng)力(li)(li)建(jian)設(she)，加(jia)大多智(zhi)(zhi)能(neng)體協同、可(ke)解釋(shi)(shi)性、模型(xing)輕量化(hua)推理等(deng)技(ji)術的(de)研(yan)究(jiu)，持續深化(hua)機器(qi)視覺、多模態、時序(xu)預測等(deng)人工智(zhi)(zhi)能(neng)關鍵技(ji)術在(zai)能(neng)源(yuan)(yuan)領域(yu)的(de)應(ying)(ying)用(yong)研(yan)究(jiu)，推動人工智(zhi)(zhi)能(neng)與能(neng)源(yuan)(yuan)領域(yu)軟件深度融合。針對人工智(zhi)(zhi)能(neng)計(ji)算(suan)耗能(neng)問題，加(jia)快突破(po)人工智(zhi)(zhi)能(neng)綠(lv)色低(di)碳技(ji)術瓶頸，研(yan)究(jiu)柔性直流供電(dian)、模塊化(hua)小型(xing)堆等(deng)能(neng)源(yuan)(yuan)供給(gei)技(ji)術，鼓勵數據中(zhong)心液(ye)冷技(ji)術、廢熱回(hui)收、備電(dian)集約(yue)化(hua)等(deng)高效能(neng)源(yuan)(yuan)綜合利(li)用(yong)技(ji)術的(de)應(ying)(ying)用(yong)。

四、保障措施

（一）強化(hua)組織實施。各(ge)地(di)(di)方能(neng)源主管部門和相關中央企(qi)業要(yao)根據意見(jian)要(yao)求，建立健(jian)全(quan)(quan)工(gong)(gong)作機(ji)制，統籌銜接好相關規劃，結合(he)實際加快推動(dong)本地(di)(di)區(qu)、本單(dan)位“人(ren)工(gong)(gong)智能(neng)+”能(neng)源的發展(zhan)，做(zuo)好各(ge)項要(yao)素保障，探索構建安(an)全(quan)(quan)治理體系，形成(cheng)上(shang)下(xia)聯動(dong)、層層落實、安(an)全(quan)(quan)發展(zhan)的工(gong)(gong)作格局，加快推進人(ren)工(gong)(gong)智能(neng)在能(neng)源領域(yu)融(rong)合(he)應(ying)用(yong)的技術研發、示范(fan)試驗、推廣應(ying)用(yong)等工(gong)(gong)作。

（二）推動(dong)協同(tong)創(chuang)(chuang)新(xin)。圍繞能源領域人工智能融合創(chuang)(chuang)新(xin)應(ying)用(yong)關鍵共(gong)性技術和配套(tao)專(zhuan)用(yong)技術，推動(dong)建(jian)設(she)一批(pi)行業(ye)研(yan)發創(chuang)(chuang)新(xin)平(ping)臺(tai)。鼓勵企業(ye)牽(qian)頭聯合科(ke)研(yan)機構(gou)、高校、社會服(fu)務機構(gou)等單位(wei)，建(jian)設(she)以技術創(chuang)(chuang)新(xin)融合應(ying)用(yong)為目標的(de)(de)跨領域、跨學科(ke)的(de)(de)“人工智能+”能源創(chuang)(chuang)新(xin)聯盟，深化(hua)產學研(yan)用(yong)合作，構(gou)建(jian)開放協同(tong)、共(gong)創(chuang)(chuang)共(gong)享(xiang)的(de)(de)能源智能化(hua)創(chuang)(chuang)新(xin)生態體系。

（三(san)）加強(qiang)標準(zhun)規(gui)(gui)范(fan)建(jian)設(she)(she)(she)。在(zai)深入總結應用(yong)(yong)示范(fan)實踐的基礎(chu)上，加快(kuai)編制能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)數據治理(li)、多元異構算(suan)力(li)融合、典型場(chang)景設(she)(she)(she)計(ji)等(deng)一(yi)批技(ji)術標準(zhun)規(gui)(gui)范(fan)，推動能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)領域(yu)人(ren)(ren)工(gong)(gong)(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)標準(zhun)體(ti)系(xi)建(jian)設(she)(she)(she)，探索建(jian)立人(ren)(ren)工(gong)(gong)(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)應用(yong)(yong)評(ping)估指標體(ti)系(xi)和行業級人(ren)(ren)工(gong)(gong)(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)應用(yong)(yong)標準(zhun)測(ce)試(shi)平臺(tai)，提升(sheng)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)領域(yu)人(ren)(ren)工(gong)(gong)(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)技(ji)術安全應用(yong)(yong)水平。鼓(gu)勵能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)企(qi)業主導制定國際標準(zhun)，以技(ji)術標準(zhun)“走出去”帶動人(ren)(ren)工(gong)(gong)(gong)智(zhi)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)技(ji)術和產品在(zai)海(hai)外(wai)能(neng)(neng)(neng)(neng)(neng)(neng)(neng)源(yuan)市場(chang)推廣應用(yong)(yong)。

（四）開(kai)(kai)展(zhan)(zhan)試(shi)點示(shi)范(fan)。組織開(kai)(kai)展(zhan)(zhan)能(neng)(neng)源(yuan)領(ling)(ling)域(yu)人(ren)(ren)工(gong)智能(neng)(neng)應(ying)(ying)用(yong)試(shi)點示(shi)范(fan)，遴選一批(pi)可復制(zhi)、易推廣的場景和(he)企業標桿應(ying)(ying)用(yong)。鼓(gu)勵(li)開(kai)(kai)展(zhan)(zhan)能(neng)(neng)源(yuan)和(he)交(jiao)通(tong)融合、油(you)氣和(he)新(xin)能(neng)(neng)源(yuan)融合等(deng)跨領(ling)(ling)域(yu)、跨行業典型場景示(shi)范(fan)。能(neng)(neng)源(yuan)領(ling)(ling)域(yu)人(ren)(ren)工(gong)智能(neng)(neng)應(ying)(ying)用(yong)相關(guan)技術(shu)裝備優先納(na)入能(neng)(neng)源(yuan)領(ling)(ling)域(yu)首臺（套）重大技術(shu)裝備支(zhi)持(chi)范(fan)圍。支(zhi)持(chi)具備條件的地區和(he)企業，因地制(zhi)宜開(kai)(kai)展(zhan)(zhan)能(neng)(neng)源(yuan)領(ling)(ling)域(yu)各(ge)類(lei)人(ren)(ren)工(gong)智能(neng)(neng)應(ying)(ying)用(yong)試(shi)點示(shi)范(fan)，在技術(shu)創新(xin)、商業模式、發展(zhan)(zhan)業態、體制(zhi)機制(zhi)等(deng)方面深入探索、先行先試(shi)。

（五(wu)）加大(da)支持(chi)力(li)度。充分發(fa)(fa)揮中央財政資金帶動作用，依托(tuo)能源領域(yu)、人(ren)(ren)工(gong)智能領域(yu)國家科(ke)技(ji)重(zhong)大(da)專(zhuan)項(xiang)和重(zhong)點研發(fa)(fa)計(ji)劃等科(ke)技(ji)專(zhuan)項(xiang)，有序推動能源領域(yu)人(ren)(ren)工(gong)智能技(ji)術應用創新。發(fa)(fa)揮多層次資本市場(chang)支持(chi)科(ke)技(ji)創新關鍵樞紐(niu)作用，引導(dao)社(she)會資本參(can)與人(ren)(ren)工(gong)智能科(ke)技(ji)項(xiang)目實施和成果(guo)轉化應用。

（六(liu)）完善(shan)人才培育生態。鼓勵能源企業與高等院校、科(ke)研院所(suo)共建“人工(gong)智能+”能源人才培養基地(di)，以行業需求為(wei)導(dao)向設(she)計跨學科(ke)課程(cheng)體(ti)系，重(zhong)點培養具備能源系統知識、人工(gong)智能算法應用(yong)能力的復(fu)合(he)(he)型(xing)人才，通過產教協同(tong)增加(jia)復(fu)合(he)(he)型(xing)人才供(gong)給。

國家(jia)發展改革委 國家(jia)能源局

2025年9月4日