非参数条件概率预测提高风电消纳的优化方法

发布时间：2018-09-10 13:06

【摘要】：功率点值预测信息未充分考虑可再生能源的不确定性及波动性,在日前发电计划和备用决策安排中存在保守或冒进的风险。引入非参数化方法,根据历史风资源状况得到风电出力的非参数条件概率预测结果,建立风电日前消纳调度模型。该模型在不同资源主导因素所划分的条件空间子集中得到预测功率条件概率分布,能够反映弃风和失负荷风险,可以根据调度需求选取不同的预测功率置信区间,计及风电不确定性完整的条件概率信息。仿真算例验证了选取合适置信水平的风电非参数化条件概率预测方法在日前优化运行中有效提高风电消纳水平,并降低系统运行风险。
[Abstract]:The prediction information of power point values does not fully consider the uncertainty and volatility of renewable energy, and there is a risk of conservatism or rash advance in the pre-day generation planning and standby decision-making arrangements. According to the historical wind resource condition, the nonparametric conditional probability prediction results of wind power output are obtained by introducing the nonparametric method, and the wind power rejection dispatching model before the day is established. In this model, the probabilistic distribution of predictive power conditions is obtained in the subset of conditional space divided by different resource dominant factors, which can reflect the risk of abandonment and loss of load, and can select different predictive power confidence intervals according to scheduling demand. Take into account the complete conditional probability information of wind power uncertainty. The simulation results show that the non-parametric conditional probabilistic prediction method of wind power with appropriate confidence level can effectively improve the wind power absorption level and reduce the operational risk of the wind power system in the pre-day optimal operation.
【作者单位】：电力系统及发电设备控制和仿真国家重点实验室(清华大学电机系);新能源与储能运行控制国家重点实验室(中国电力科学研究院);
【基金】：国家重点研发计划支持项目(2016YFB0900105) 国家电网公司科技项目(XTB17201500037)~~
【分类号】：TM614;TM73

【参考文献】