電力營(yíng)銷領(lǐng)域中大語言模型的多輪對(duì)話技術(shù)的研究
王江輝1，劉鵬飛2
(1 江蘇方天電力技術(shù)有限公司，江蘇 南京 211102；
2 江蘇思極科技服務(wù)有限公司，江蘇 南京 210024)
摘 要：電力營(yíng)銷領(lǐng)域的多輪對(duì)話技術(shù)，旨在提高電力營(yíng)銷效率和用戶滿意度。系統(tǒng)梳理了多輪對(duì)話技術(shù)的相關(guān)理論，闡述了電力營(yíng)銷多輪交互的預(yù)處理模塊總體設(shè)計(jì)，介紹了數(shù)據(jù)預(yù)處理模塊的功能及其實(shí)現(xiàn)方法。利用這些模塊，在電力營(yíng)銷領(lǐng)域進(jìn)行了多輪實(shí)證研究，分析了系統(tǒng)實(shí)驗(yàn)結(jié)果，并總結(jié)了多輪對(duì)話技術(shù)在電力行業(yè)的優(yōu)勢(shì)，展望了未來的研究方向，希望對(duì)電力營(yíng)銷領(lǐng)域的多輪對(duì)話技術(shù)的發(fā)展提供借鑒。
關(guān)鍵詞： 多輪對(duì)話技術(shù)；電力營(yíng)銷；預(yù)處理
中圖分類號(hào)：F407.61 ；TP18 文獻(xiàn)標(biāo)識(shí)碼：A 文章編號(hào)：1007-3175(2025)01-0072-05
Research on Multi-Round Dialogue Technology of Large Language
Models in the Field of Power Marketing
WANG Jiang-hui1, LIU Peng-fei2
（1 Jiangsu Frontier Electric Technology Co., Ltd, Nanjing 211102, China;
2 Jiangsu Siji Technology Co., Ltd, Nanjing 210024, China）
Abstract: The multi-round dialogue technology in the field of power marketing aims to improve the efficiency of power marketing and user satisfaction. The relevant theories of multi-round dialogue technology are systematically sorted out, the overall design of the preprocessing module for multi-round interaction in power marketing is elaborated, and the functions of the data preprocessing module and its realization methods are introduced. Using these modules, a multi-round empirical study was conducted in the field of electric power marketing,the results of system experiments were analyzed, and the advantages of multi-round dialogue technology in the electric power industry were summarized, and the future research direction was looked forward to, hoping to provide a reference for the development of multi-round dialogue technology in the field of electric power marketing.
Key words: multi-round dialogue technology; power marketing; preprocessing
參考文獻(xiàn)
[1] 曹祎，張莉，郭靜，等. 基于大語言模型的低碳電力市場(chǎng)發(fā)展應(yīng)用前景[J] . 智慧電力，2024，52(2) ：8-16.
[2] 于碩，王司宇，王超，等. 類 ChatGPT 大語言模型在電力系統(tǒng)中的應(yīng)用前景[J] . 電氣時(shí)代，2023(10) ：50-53.
[3] 馮鈞，暢陽紅，陸佳民，等. 基于大語言模型的水工程調(diào)度知識(shí)圖譜的構(gòu)建與應(yīng)用[J] . 計(jì)算機(jī)科學(xué)與探索，2024，18(6) ：1637-1647.
[4] 趙俊華，文福拴，黃建偉，等. 基于大語言模型的電力系統(tǒng)通用人工智能展望：理論與應(yīng)用[J]. 電力系統(tǒng)自動(dòng)化，2024，48(6) ：13-28.
[5] 谷奉錦. 基于知識(shí)圖譜的 5G 網(wǎng)絡(luò)故障診斷與分析方法研究[D]. 南京：南京郵電大學(xué)，2023.
[6] 潘雨黛，張玲玲，蔡忠閩，等. 基于大規(guī)模語言模型的知識(shí)圖譜可微規(guī)則抽取[J] . 計(jì)算機(jī)科學(xué)與探索，2023，17(10) ：2403-2412.
