基于相依关系的新能源功率预测场景生成及调度应用

新能源日前功率预测对指导电网计划编制具有重要意义，但目前预测水平制约了预测结果的充分应用，为此文中提出新能源功率预测的场景生成方法和预测结果纳入调度计划编制的思路。首先，通过对比分析新能源发电功率和预测功率的边缘分布，建立了揭示二者相依结构的Copula模型，并提出了基于相依关系的多预测场景建模方法。随后，通过分析预测偏差对供电平衡的影响，利用多场景下不同时段预测偏差的规律，提出了将预测可信度纳入省级电网调度计划编制的方法。相关成果已经在国家电网公司西北电力调控分中心得到实际应用，在新能源纳入备用的基础上，采取优化常规电源运行方式、合理组织交易互济等措施，2017年增发新能源电量为4.7 TW·h，降低了3%的受阻率，为新能源纳入电网调度计划编制提供了依据。

Dependence Relationship Based Scenario Generation of Power Prediction for Renewable Energy and Its Application in Dispatch

The day-ahead forecasting of renewable energy is significant for guiding preparation of power grid planning, but the current prediction accuracy restricts the full application of prediction results. The idea of scenario generation method for renewable energy prediction and prediction consequence which is added into dispatch plan is proposed. Firstly, by comparing and analyzing marginal distribution of generated power and predictive power for renewable energy, a Copula model is established to reveal the dependence relationship between two distributions. Meanwhile, the multi-prediction scenario simulation method based on the dependence relationship is presented. Secondly, the method which integrates the reliability of prediction into provincial power grid dispatch plan is put forward by analyzing the impact of prediction deviation on power supply balance and the law of prediction deviations in multiple scenarios. The relevant results have been applied in the northwest power control sub-center of State Grid Corporation of China. On the basis of the reserve for renewable energy, the measurements which utilizes optimization of operation for conventional power resources, organizes transaction of power market are taken. In 2017, the renewable energy generating capacity is increased by 4. 7 GW·h and the blocking rate is reduced by 3%, which provides a potent support of incorporating renewable energy into the power grid dispatching plan.