《電子技術(shù)應(yīng)用》
您所在的位置:首頁 > 模擬設(shè)計(jì) > 設(shè)計(jì)應(yīng)用 > 電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目自適應(yīng)檢索方法研究
電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目自適應(yīng)檢索方法研究
電子技術(shù)應(yīng)用
馬玉龍,俞陽,王逸民,王婭
國(guó)網(wǎng)江蘇省電力有限公司營(yíng)銷服務(wù)中心
摘要: 針對(duì)電力系統(tǒng)數(shù)據(jù)具有復(fù)雜性、多樣性特點(diǎn),導(dǎo)致檢索難度過高的問題,設(shè)計(jì)了電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目自適應(yīng)檢索方法。依據(jù)電力系統(tǒng)出力變化率,選取二分量一維混合高斯模型,構(gòu)建電力系統(tǒng)波動(dòng)概率分布模型。對(duì)比概率分布模型模擬的電力系統(tǒng)波動(dòng)數(shù)據(jù)與量測(cè)數(shù)據(jù),依據(jù)判定閾值辨識(shí)電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目,構(gòu)建數(shù)據(jù)條目檢索庫(kù)。利用哈希函數(shù)獲取檢索庫(kù)內(nèi)較大波動(dòng)數(shù)據(jù)條目的哈希特征,生成二值碼。較大波動(dòng)數(shù)據(jù)條目檢索時(shí),生成用戶檢索詞的二值編碼,計(jì)算檢索詞二值碼與檢索庫(kù)內(nèi)條目二值碼的漢明距離,并對(duì)其加權(quán)處理,利用加權(quán)漢明距離排序數(shù)據(jù)條目,獲取較大波動(dòng)數(shù)據(jù)條目的自適應(yīng)檢索結(jié)果。實(shí)驗(yàn)結(jié)果表明,該方法能夠依據(jù)用戶輸入的檢索詞,自適應(yīng)檢索電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目,檢索結(jié)果的歸一化折損累積增益均高于0.9,檢索時(shí)間低于500 ms。
中圖分類號(hào):TM732 文獻(xiàn)標(biāo)志碼:A DOI: 10.16157/j.issn.0258-7998.256297
中文引用格式: 馬玉龍,俞陽,王逸民,等. 電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目自適應(yīng)檢索方法研究[J]. 電子技術(shù)應(yīng)用,2025,51(6):27-31.
英文引用格式: Ma Yulong,Yu Yang,Wang Yimin,et al. Research on adaptive retrieval method for data entries with large fluctuations in power system[J]. Application of Electronic Technique,2025,51(6):27-31.
Research on adaptive retrieval method for data entries with large fluctuations in power system
Ma Yulong,Yu Yang,Wang Yimin,Wang Ya
Marketing Service Center, State Grid Jiangsu Electric Power Co., Ltd.
Abstract: Aiming at the problem of high retrieval difficulty caused by the complexity and diversity of power system data, this paper studies an adaptive retrieval method for large fluctuation data entries in the power system. Based on the rate of change in power system output, a two-component one-dimensional mixture Gaussian model is selected to construct a probability distribution model for power system fluctuations. Compare the power system fluctuation data simulated by the probability distribution model with the measurement data, identify the large fluctuation data entries in the power system based on the judgment threshold, and construct a data entry retrieval library. Use hash functions to obtain hash features of large fluctuation data entries in the retrieval database and generate binary codes. When retrieving large fluctuation data entries, generate binary codes for user search terms, calculate the Hamming distance between the binary codes of search terms and the binary codes of entries in the search database, and weight them. Use the weighted Hamming distance to sort the data entries and obtain adaptive search results for large fluctuation data entries. The experimental results show that this method can adaptively retrieve large fluctuation data entries in the power system based on user input search terms. The normalized cumulative loss gain of the retrieval results is higher than 0.9, and the retrieval time is less than 500 ms.
Key words : power system;large fluctuations;data entries;adaptive;search method;Hanming distance

引言

電力系統(tǒng)產(chǎn)生的數(shù)據(jù),涵蓋了從發(fā)電、輸電、配電到用電等各個(gè)環(huán)節(jié)的實(shí)時(shí)數(shù)據(jù)、歷史數(shù)據(jù)以及各類事件記錄[1]。數(shù)據(jù)的采集、存儲(chǔ)和分析已成為保障電網(wǎng)穩(wěn)定運(yùn)行和高效管理的重要基石。較大波動(dòng)數(shù)據(jù)不局限于單一負(fù)荷數(shù)據(jù),而是涵蓋電力系統(tǒng)中多個(gè)負(fù)荷數(shù)據(jù),通過對(duì)系統(tǒng)整體數(shù)據(jù)的分析來確定。這是因?yàn)殡娏ο到y(tǒng)各負(fù)荷之間相互關(guān)聯(lián),多個(gè)負(fù)荷的波動(dòng)情況對(duì)于全面評(píng)估系統(tǒng)運(yùn)行狀態(tài)至關(guān)重要。電力系統(tǒng)較大波動(dòng)的數(shù)據(jù)條目往往蘊(yùn)含著重要的信息,如設(shè)備故障預(yù)警、負(fù)荷突變、電網(wǎng)穩(wěn)定性問題等,對(duì)于電力系統(tǒng)的監(jiān)控、分析和優(yōu)化至關(guān)重要[2]。傳統(tǒng)的數(shù)據(jù)檢索方法往往依賴于固定的查詢條件或索引結(jié)構(gòu),難以適應(yīng)電力系統(tǒng)數(shù)據(jù)動(dòng)態(tài)變化、多維度、高復(fù)雜性的特點(diǎn)。研究能夠自適應(yīng)電力系統(tǒng)數(shù)據(jù)波動(dòng)特性的檢索方法能夠提升電力系統(tǒng)的智能化水平。

趙松燕等[3]提出MapReduce能夠在大規(guī)模的分布式計(jì)算集群中處理大量的數(shù)據(jù),根據(jù)數(shù)據(jù)量的增加而自動(dòng)擴(kuò)展,支持處理超大規(guī)模的數(shù)據(jù)集。在輸電監(jiān)測(cè)數(shù)據(jù)智能檢索中,該模型能夠處理電流、電壓、溫度等多樣化的監(jiān)測(cè)數(shù)據(jù)類型。在特定應(yīng)用場(chǎng)景下,能夠?qū)崿F(xiàn)接近實(shí)時(shí)的數(shù)據(jù)處理和檢索,適用于需要快速響應(yīng)的輸電監(jiān)測(cè)場(chǎng)景。但是MapReduce在處理大數(shù)據(jù)集時(shí),需要將數(shù)據(jù)從輸入節(jié)點(diǎn)傳輸?shù)綀?zhí)行Map和Reduce函數(shù)的節(jié)點(diǎn),導(dǎo)致大量的數(shù)據(jù)移動(dòng)。數(shù)據(jù)移動(dòng)過程將消耗大量的時(shí)間和網(wǎng)絡(luò)帶寬,成為性能瓶頸。趙征宇等[4]通過融合多個(gè)語義粒度的語義信息,處理更復(fù)雜的語義信息,包括短語、句子甚至段落級(jí)別的語義匹配,從而增強(qiáng)對(duì)文本內(nèi)容的理解能力,有助于在檢索過程中更準(zhǔn)確地匹配用戶的查詢意圖和文檔內(nèi)容,提高檢索滿意度。通過調(diào)整語義粒度的選擇和融合策略,針對(duì)不同領(lǐng)域的文本特點(diǎn)和用戶需求進(jìn)行優(yōu)化。但是某些粒度的語義信息非常稀疏,導(dǎo)致難以進(jìn)行有效地匹配和檢索。劉東等人研究電網(wǎng)調(diào)控信息的智能檢索方法[5]。通過構(gòu)建電網(wǎng)設(shè)備、屬性、關(guān)聯(lián)關(guān)系等本體概念框架,實(shí)現(xiàn)對(duì)電網(wǎng)調(diào)控信息的層次化、結(jié)構(gòu)化表示。知識(shí)圖譜能夠捕捉文本中的語義關(guān)系,如設(shè)備間的關(guān)聯(lián)、屬性間的依賴等,有助于系統(tǒng)更準(zhǔn)確地理解用戶的查詢意圖,從而提供更精確的檢索結(jié)果。通過知識(shí)圖譜,系統(tǒng)可以自動(dòng)感知電網(wǎng)事件,如故障異常、倒閘操作等,將電網(wǎng)調(diào)控信息以結(jié)構(gòu)化的形式存儲(chǔ),便于不同部門之間的知識(shí)共享與復(fù)用。但是知識(shí)圖譜的構(gòu)建需要收集大量的電網(wǎng)調(diào)控信息,并進(jìn)行清洗、標(biāo)注和融合等處理。且電網(wǎng)調(diào)控信息具有動(dòng)態(tài)性和時(shí)效性,需要不斷更新和維護(hù)。邸劍等研究利用BERT和覆蓋率機(jī)制改進(jìn)的HiNT文本檢索模型[6],該模型具有強(qiáng)大的語義表示能力,能夠?qū)ξ臋n的各個(gè)段落進(jìn)行深度語義理解,使模型在檢索過程中,能夠更準(zhǔn)確地匹配用戶的查詢意圖和文檔內(nèi)容。但是該模型過于依賴于預(yù)訓(xùn)練的語言模型或語義表示模型來提取文本的語義信息。如果預(yù)訓(xùn)練模型的質(zhì)量不高或與實(shí)際應(yīng)用場(chǎng)景不匹配,直接影響檢索性能。研究電力系統(tǒng)較大波動(dòng)數(shù)據(jù)條目的自適應(yīng)檢索方法,實(shí)現(xiàn)電力系統(tǒng)波動(dòng)數(shù)據(jù)的精準(zhǔn)捕捉。該模型能夠自動(dòng)學(xué)習(xí)電力數(shù)據(jù)的波動(dòng)模式,根據(jù)數(shù)據(jù)的實(shí)時(shí)變化調(diào)整檢索參數(shù),從而提高檢索準(zhǔn)確性。


本文詳細(xì)內(nèi)容請(qǐng)下載:

http://theprogrammingfactory.com/resource/share/2000006558


作者信息:

馬玉龍,俞陽,王逸民,王婭

(國(guó)網(wǎng)江蘇省電力有限公司營(yíng)銷服務(wù)中心,江蘇 南京 210024)


Magazine.Subscription.jpg

此內(nèi)容為AET網(wǎng)站原創(chuàng),未經(jīng)授權(quán)禁止轉(zhuǎn)載。