| 引用本文: | 赵书强,赵蓬飞,韦子瑜,廖艺茗,王子巍.数据驱动下考虑多预测误差带信息的多场景随机优化调度[J].电力自动化设备,2024,44(11):52-59. |
| ZHAO Shuqiang,ZHAO Pengfei,WEI Ziyu,LIAO Yiming,WANG Ziwei.Multi-scenario stochastic optimal scheduling considering multi-prediction error band information under data-driven[J].Electric Power Automation Equipment,2024,44(11):52-59. |
|
| 摘要: |
| 针对高比例可再生能源接入电网后省级电网面临的风光功率预测不确定性大、调度决策难等问题，提出一种多场景随机优化调度模型。基于核密度估计方法引入调度预测误差带的概念，结合风光聚合场站的预测出力和实际出力构建计及时间相关性的海量随机场景，结合改进的K-means聚类与同步回代消除算法进行场景缩减，引入布莱尔分数评价削减后随机场景的性能；基于风光随机出力场景，考虑供热系统中的热用户舒适度，构建日前随机优化调度模型。基于新疆电网实际网架结构和风光实际历史出力数据进行算例验证，结果表明：所提方法生成的风光出力随机场景可以更好地描述风光的预测轨迹；相较于传统日前调度方法，基于场景法的随机优化调度模型可有效提升风光可再生能源消纳率；以供热系统中热用户舒适度为导向，热电联产机组参与热力系统热电调节可进一步降低风光弃电率。 |
| 关键词: 风电 光伏发电 预测误差带 场景生成 随机优化调度 |
| DOI：10.16081/j.epae.202312036 |
| 分类号:TM73 |
| 基金项目: |
|
| Multi-scenario stochastic optimal scheduling considering multi-prediction error band information under data-driven |
|
ZHAO Shuqiang, ZHAO Pengfei, WEI Ziyu, LIAO Yiming, WANG Ziwei
|
|
Key Laboratory of Distributed Energy Storage and Microgrid of Hebei Province, North China Electrical Power University, Baoding 071003, China
|
| Abstract: |
| Aiming at the problems such as large uncertainty of wind and photovoltaic prediction and difficult scheduling and decision-making faced by provincial power grid after a high proportion of renewable energy is connected to the power grid, a multi-scenario stochastic optimal scheduling model is proposed. The concept of scheduling prediction error band is introduced based on the kernel density estimation method. Combining with the predicted output and actual output of the wind and photovoltaic aggregated power station, massive random scenarios are constructed considering the time dependence. The improved K-means clustering and synchronous back elimination algorithm are combined for scenario reduction, and the Brier score is introduced to evaluate the performance of the reduced random scenario. Based on the wind and photovoltaic random output scenario, a day-ahead random optimal scheduling model is constructed considering the comfort of thermal users in the heating system. Example verification is carried out based on the actual grid structure of Xinjiang Power Grid and the actual historical wind and photovoltaic output data, and the results show that the wind and photovoltaic output random scenarios generated by the proposed method can better describe the prediction track of wind and photovoltaic, the stochastic optimal scheduling model based on the scenario method can effectively improve the accommodation rate of wind and photovoltaic renewable energy compared with the traditional day-ahead scheduling method. Guided by the comfort of thermal users in the heating system, combined heat and power units participating in the thermoelectric regulation of thermal system can further reduce the wind and photovoltaic power abandonment rate. |
| Key words: wind power photovoltaic power generation prediction error band scenario generation stochastic optimal scheduling |
