论文部分内容阅读
风功率预测误差的高精度拟合与置信水平的科学选取是保障风电高效参与日前机组组合(UC)的关键问题，该文将在同时考虑上述两方面问题基础上，研究含风电的UC决策模型。首先，基于误差特性分析提出一种时序下的预测误差分段拟合方法，利用t location-scale分布进行拟合，以改善厚尾效应，提高拟合精度，更可在时序上与UC决策相配合；其次，构建了可以同时考虑传统成本、额外备用成本与风险成本的双分位点型UC决策模型，通过不同成本间的制约关系平衡置信水平的选取，通过不同置信水平的划分指导备用分类，通过时变置信水平适应误差时序分段分布，以此使模型更具经济性、针对性与适用性；最后，采用带有启发式搜索原则的改进混合粒子群算法，求解文中的多变量混合整数规划模型，算例结果验证了所提方法的有效性。“,”There are two problems in unit commitment (UC) with wind power integration, high fitting precision of forecast error and appropriate selection of confidence interval. Aiming to solve the above problems, this paper proposed a new UC model with wind power integration. The first, based on the analysis of forecast error, a temporal segment method of forecast error was proposed, which usedt location-scale distribution for fat tail effect, improving the fitting accuracy, and also coordinated with day-ahead UC in time series. The second, a new UC model with double quantiles was proposed, which considered traditional coasts, extra reserve costs and risk costs. In this model, the confidence interval determined by the constrains among different costs, the reserve style sorted by dividing different confidence interval, the time-varying confidence level used to correspond to temporal segment distribution, make the model more applicable and economic. The last, the improved hybrid particle swarm optimization algorithm with heuristic searching strategy was proposed to solve the multivariate mixed integer programming problem, and the simulation results demonstrate the effectiveness of the proposed model.