中国电力 ›› 2025, Vol. 58 ›› Issue (8): 130-138.DOI: 10.11930/j.issn.1004-9649.202311142

• 新型电网 • 上一篇    下一篇

充电用能约束下工业园区负荷管控的随机优化策略

茹传红1(), 卢姬1(), 秦建1(), 张军达2, 常俊晓1, 蒋贝妮1   

  1. 1. 国网浙江台州供电公司,浙江 台州 318000
    2. 国网浙江综合能源服务有限公司,浙江 杭州 310016
  • 收稿日期:2023-11-30 发布日期:2025-08-26 出版日期:2025-08-28
  • 作者简介:
    茹传红(1972),男,高级工程师,硕士,从事电力系统生产与管理研究,E-mail:13957563868@139.com
    卢姬(1990),女,通信作者,工程师,硕士,从事综合能源服务策略研究,E-mail:980749437@qq.com
    秦建(1974),男,经济师,从事电力营销管理研究,E-mail:879010308@qq.com
  • 基金资助:
    国网浙江省电力有限公司科技项目(5211TZ230002);国家自然科学基金资助项目(52107122)。

Stochastic Optimization Strategy for Load Management of Industrial Park Under Energy Constraints

RU Chuanhong1(), LU Ji1(), QIN Jian1(), ZHANG Junda2, CHANG Junxiao1, JIANG Beini1   

  1. 1. State Grid Taizhou Power Supply Company, Taizhou 318000, China
    2. State Grid Zhejiang Integrated Energy Service Company, Hangzhou 310016, China
  • Received:2023-11-30 Online:2025-08-26 Published:2025-08-28
  • Supported by:
    This work is supported by Science and Technology Project of State Grid Zhejiang Electric Power Co., Ltd. (No.5211TZ230002) & National Natural Science Foundation of China (No.52107122).

摘要:

为解决工业园区微电网在能源短缺情况下的负荷管理问题,提出一种管理用户特定负荷的新方法。首先,将用户特定负荷管理定义为一个随机预测模型控制问题,对光伏和用户电力需求进行预测建模。然后,结合两阶段随机规划和近似动态规划对其进行求解。最后,在模拟用户对负荷管理响应的环境中,通过2种控制器测试了替代解决方案的有效性。结果表明,即使没有完整的用户响应模型,控制器使用预测模型进行负荷管理也可以显著提高电力可用性和用户的消费效益。

关键词: 工业园区, 负荷管理, 随机预测模型, 两阶段随机规划, 近似动态规划

Abstract:

In order to solve the load management problem of industrial park microgrids in the event of energy shortage, this paper proposes a new method for managing user-specific loads. By defining the user-specific load management as a stochastic predictive model control problem, this method firstly establishes prediction models for photovoltaic and users’ power demand, and then solve them using two-stage stochastic programming and approximate dynamic programming. Finally, the effectiveness of the alternative solutions is tested through two controllers in an environment in which user response to the load management is simulated. The results indicate that even if the controller does not have a complete model of user response, the predictive models used for scheduling the load management can significantly improve the power availability and user consumption efficiency.

Key words: industrial park, load management, random prediction model, two stage stochastic programming, approximate dynamic programming


AI


AI小编
您好!我是《中国电力》AI小编,有什么可以帮您的吗?