| 14 | 0 | 37 |
| 下载次数 | 被引频次 | 阅读次数 |
针对电网企业代理购电系统在现货市场条件下各类型交易品种组合购电策略、量价预测方面存在的局限，从代理购电业务的发展出发，提出了基于数据驱动的代理购电交易数据回溯分析方法，在交易复盘的基础上利用多维属性增强的深度强化学习完成交易价格申报优化重构。此外，考虑突发疫情、季节变化等随机因素，利用伪量测训练方法对样本数据进行分析复盘，提出基于深度神经网络技术的代理购电交易优化决策方案。今后除明确代理用户的用能特性外，还应加强相应仿真推演平台的建设，以精准分析市场交易过程。
Abstract：In view of the limitations of power purchase strategy and volume price prediction of various types of trading varieties under the spot market conditions, starting from the development of agency power purchase business, a data-driven retrospective analysis method of agency power purchase transaction data is proposed, and the multi-dimensional attribute enhanced deep reinforcement learning is used to complete the optimization and reconstruction of transaction price declaration on the basis of transaction review. In addition, considering random factors such as sudden epidemics and seasonal changes, the pseudo-measurement training method is used to analyze and review the sample data, and an optimized decision-making scheme for agency power purchase transactions based on deep neural network technology is proposed. In the future, except for clarifying the energy consumption characteristics of proxy users, the construction of corresponding simulation and deduction platforms should be strengthened to accurately analyze the market trading process.
[1] 中国政府网.国家发展改革委关于进一步深化燃煤发电上网电价市场化改革的通知[EB/OL].(2021-10-11)[2022-10-12] http://www.gov.cn/zhengce/zhengceku/2021-10/12/content_5642159.htm.
[2] 中国政府网.国家发展改革委办公厅关于组织开展电网企业代理购电工作有关事项的通知[EB/OL].(2021-10-23)[2022-10-12] http://www.gov.cn/zhengce/zhengceku/2021-10/27/content_5645848.htm.
[3] 国家能源局.国家能源局2022年深化"放管服" 改革优化营商环境重点任务分工方案[EB/OL].(2022-08-25)[2022-11-23].http://www.nea.gov.cn/2022-08/25/c_1310655795.htm.
[4] 国家发展与改革委员会.国家发展改革委办公厅、国家能源局综合司关于开展电力现货市场建设试点工作的通知[EB/OL]. (2017-8-28)[2022-11-24].https://www.ndrc.gov.cn/xxgk/zcfb/tz/201709/t20170905_962552.html?code=&state=123.
[5] 国家发展与改革委员会.关于加快建设全国统一电力市场体系的指导意见[EB/OL].(2022-01-18)[2022-12-01].http://www.gov.cn/zhengce/zhengceku/2022-01/30/content_5671296.htm.
[6] 陈曦.解析电网企业代理购电:对电价、电网、售电市场影响几何[EB/OL].(2021-10-27)[2022-11-30].https://www.thepaper.cn/newsDe-tail_forward_15091165.
[7] 解勇, 徐伟栋, 周奕辰.代理购电机制出台对电网企业及市场的影响研究[J].电力设备管理, 2022(10):233-235. XIE Yong, XU Weidong, ZHOU Yichen. Study on the Influence of the Introduction of Agent Purchasing Mechanism on Power Grid Enterprises and Market[J]. Electric Power Equipment Management, 2022(10):233-235.
[8] 史岳, 王江波, 钟心蕊.现货环境下的电网代理购电业务[J].中国电力企业管理, 2022(19):66-68. SHI Yue, WANG Jiangbo, ZHONG Xinrui. Power Network Agent Purchasing Business in Spot Environment[J]. China Power Enterprise Management, 2022(19):66-68.
[9] 温宇晴.电网代理购电参与现货结算[J].中国电力企业管理, 2022(22):40. WEN Yuqing. Power Grid Agent to Participate in Spot Settlement[J]. China Power Enterprise Management, 2022(22):40.
[10] 陈睿.电力市场代理报价策略研究及平台开发[D].南京:东南大学, 2018.
[11] 王亚微, 梁爽, 穆殿忠, 等.一种基于用电分析的智能购电提醒方法及系统:CN114037400A[P].2022-02-11.
[12] 任艺.考虑用户需求响应的售电公司购售电决策研究[D].北京:华北电力大学, 2017.
[13] 戴湘豫, 周小花, 李衡, 等.MACD与K线图技术在证券投机中的实证研究[J].湖南工业大学学报, 2010, 24(4):102-105. DAI Xiangyu, ZHOU Xiaohua, LI Heng, et al. An Empirical Study of MACD and K Chart Techniques in Securities Speculation[J]. Power System Technolgy, 2010, 24(4):102-105.
[14] 罗松山.基于改进灰度模型的中长期电力负荷预测方法[D].长沙:湖南大学, 2018.
[15] 彭文, 王金睿, 尹山青.电力市场中基于Attention-LSTM的短期负荷预测模型[J].电网技术, 2019, 43(5):1745-1751. PENG Wen, WANG Jinrui, YIN Shanqing. Short-term Predictive Load Model Based on Attention-LSTM in Electricity Market[J]. Power System Technolgy, 2019, 43(5):1745-1751.
[16] 刘俊宇, 陈辉, 张福峰, 等.基于改进随机森林算法的电力企业市场竞争预测模型构建与实现[J].机械设计与制造工程, 2021, 50(6):85-88. Liu Junyu, Chen Hui, Zhang Fufeng, et al. Construction and implementation of power enterprise market competition prediction model based on improved stochastic forest algorithm[J]. Mechanical Design and Manufacturing Engineering, 2021, 50(6):85-88.
[17] 陈靖邦, 潘俊哲, 沈皓朗, 等.一种多趋势指标结合与择时引入峰值的投资组合优化系统[J].计算机科学, 2021, 48(增刊2):693-698. CHEN Jingbang, PAN Junzhe, SHEN Haolang, et al. Portfolio Optimization System Based on Multiple Trend Indices with Time Picking of Inducing Peak Price[J]. Computer Science, 2021, 48(S2):693-698.
[18] 孙伟昊.基于Hilbert变换的船舶结构非线性横摇运动水动系数辨识方法[D].烟台:鲁东大学, 2021.
[19] 广发证券.基于GFTD模型的股票多空组合策略系统[EB/OL]. (2012-05-09)[2022-12-01].https://www.docin.com/p-423004102.html.
[20] 董竹, 周悦.股票市场的惯性效应、择时策略与交易规则设计[J] 统计与决策, 2019, 35(12):146-149. DONG Zhu, ZHOU Yue. Momentum Effect, Timing Strategy and Transaction Rules Design in Stock Market[J]. Statistics & Decision, 2019, 35(12):146-149.
[21] 周自强, 王韵楚, 颜拥, 等.面向电网企业代理购电的行业精细化电价定价机制[J/OL]. 电力系统自动化:1-14[2023-05-12]. http://kns.cnki.net/kcms/detail/32.1180.TP.20221201.1637.006.html.
基本信息:
DOI：10.19929/j.cnki.nmgdljs.2023.0039
中图分类号:
引用信息:
[1]李彬1,郭慧芳1,薛利2等.代理购电业务发展及辅助决策支撑关键技术研究[J].内蒙古电力技术,2023(03):51-56.DOI:10.19929/j.cnki.nmgdljs.2023.0039.
基金信息:
国家自然科学基金资助项目“可再生能源接入下的大规模负荷感知模型及调控策略研究”（51777068）