应用全场景可行安全约束机组组合的风电最优消纳模型及解法

An Optimal Accommodation Model of Wind Power and Its Solution Based on All-Scenario-Feasible Security Constrained Unit Commitment

针对含风电电力系统的安全约束机组组合(SCUC)问题,目前广泛采用的鲁棒优化方法在求解时不能满足调度解的非预期约束并存在严重缺陷,且各种模型和算法中均未考虑最大消纳和最优经济消纳的区别。为克服以上困难,提出了满足非预期约束条件的风电最大消纳和最优消纳模型,并给出对应求解方法。该模型及对应求解方法为容纳可能的弃风决策以扩大调度解空间,引入了风电可变不确定集来替代广泛采用的固定不确定集;为满足调度解的非预期性,引入了一组规模极小的强非预期约束,避免了传统建模方式导致的大量非预期约束难题;为克服鲁棒优化中复杂极小极大结构带来的求解困难,引入了基于可变不确定集顶点场景的全场景可行调度模型,该模型具有单层混合整数线性规划(MILP)问题结构,求解可得调度问题的最优解。实验及测试结果表明:该算法在IEEE118节点系统中取得了良好效果,显示了最大消纳和最优消纳的区别,可以实现含风电SCUC问题的快速求解,并使风电消纳的经济性得到提高。

For the security constrained unit commitment (SCUC) with wind power, the nonanticipative constraints are not satisfied by the widely adopted two-stage robust optimization based methods, and there exists a very serious deficiency. Meanwhile, the maximum accommodation and the optimal economic accommodation of the wind energy are not clearly distinguished from each other. In this study, models and solution methods for solving the maximum accommodation and the optimal economic accommodation of the wind energy in SCUC are proposed to address these issues. A variable uncertainty set of wind power is introduced to replace the widely adopted fixed uncertainty set so that the model and the method can accommodate possible wind curtailment decision to expand the scheduling solution space. A group of small scale strong nonanticipative constraints is introduced to guarantee the nonanticipativity of the scheduling solution and to avoid the huge number of nonanticipative constraints in the traditional modelling methods. An all-scenario-feasible scheduling model based on the vertices of the variable uncertainty set is established to avoid the complex min-max structure in the robust optimization based framework, so that only a single level mixed integer linear programming (MILP) problem needs to be solved in the proposed method, and the optimal scheduling solution is obtained. Numerical results on an IEEE 118-bus system show that the performance of the proposed algorithm is satisfactory. The differences between the maximum accommodation and the optimal economic accommodation of the wind energy are clearly revealed and the economic accommodation level of wind power is increased.