| 引用本文: | 吕海鹏,希望·阿不都瓦依提,孟令鹏.计及源-荷预测不确定性的微电网双级随机优化调度[J].电力自动化设备,2022,42(9): |
| Lü Haipeng,XIWANG·Abuduwayiti,MENG Lingpeng.Two-level stochastic optimal scheduling of microgrid considering uncertainty of source-load prediction[J].Electric Power Automation Equipment,2022,42(9): |
|
| 摘要: |
| 风光储微电网接入高渗透率的可再生能源对其经济运行构成了巨大的挑战。针对这一问题，提出了计及源-荷预测不确定性的微电网双级调度策略。在日前调度阶段，以多场景下的期望运行成本最低为优化目标，构建了基于多场景技术的随机优化调度模型。利用场景分析法对日前风电、光伏和负荷预测进行场景分析；建立了多场景下含不确定变量的功率平衡方程，并将其松弛为不等式后作为一个随机事件使其以较高的概率满足机会约束；此外，用机会约束规划构建了旋转备用容量的可靠性约束模型，使微电网在一定的置信水平下满足系统的可靠运行。在日内调度阶段，提出了结合自适应小波包算法的日内滚动调度模型。利用自适应小波包算法动态提取每一控制周期内超短期预测数据与日前调度计划之间的功率偏差，并由蓄电池、超级电容器和主网供电共同平抑。 |
| 关键词: 随机优化 机会约束 多场景技术 自适应小波包算法 混合储能 |
| DOI：10.16081/j.epae.202203030 |
| 分类号:TM727;TM73 |
| 基金项目:国家自然科学基金资助项目(52067021)；新疆维吾尔族自治区重点研发计划项目(2020B02001) |
|
| Two-level stochastic optimal scheduling of microgrid considering uncertainty of source-load prediction |
|
Lü Haipeng, XIWANG·Abuduwayiti, MENG Lingpeng
|
|
School of Electrical Engineering, Xinjiang University, Urumqi 830002, China
|
| Abstract: |
| The access of high permeability of renewable energy to wind-solar-storage microgrid poses a huge challenge to its economic operation. To solve this problem, the two-level scheduling strategy of microgrid considering uncertainty of source-load prediction is proposed. In the day-ahead scheduling stage, the stochastic optimal scheduling model based on multi-scenario technology is constructed taking the minimum expected operating cost in multi-scenarios as the optimization objective. The scenario analysis method is used to analyze the day-ahead prediction of wind power and photovoltaic, together with load demand. Then, the power balance equation with uncertain variables in multi-scenarios is established, and it is relaxed to inequality and used as a random event to satisfy the chance constraint with high probability. In addition, the reliability constraint model of spinning reserve capacity is constructed by chance-constrained programming, so that the microgrid can operate reliably at a certain confidence level. In the intra-day scheduling stage, an intra-day rolling scheduling model combined with an adaptive wavelet packet algorithm is proposed. The adaptive wavelet packet algorithm is used to dynamically extract the power deviation between the ultra-short-term prediction data in each control cycle and the day-ahead scheduling plan, which is suppressed by the battery, supercapacitor and main grid. |
| Key words: stochastic optimization chance constraint multi-scenario technology adaptive wavelet packet algorithm hybrid energy storage |
