电力系统恢复的主从递阶决策模型及其优化算法

引入主从递阶决策概念，建立了一种适合于大面积停电后分布式恢复的主从递阶决策模型。该模型包括主从递阶恢复决策结构和基于承诺与约定的主从递阶决策模型。通过自下而上的可行性检验与自上而下的目标优化，限制问题的求解规模，避免了对大规模系统直接求解的困难。针对主从递阶决策问题的优化求解，提出基于回溯算法的优化方法。山东电网的仿真结果表明，该方法充分考虑了电力系统分层递阶调度的特点，能满足大规模电力系统大面积停电后分层分区恢复协调调度决策的要求。

Leader-Follower Hierarchical Decision Model and Optimal Method for Power System Restoration

From the viewpoint of STACKELBERG decision problems, a leader-follower hierarchical decision model is proposed for distributed restoration after blackout. This model includes a leader-follower hierarchical restoration decision-making structure and coordination based on commitments and conventions. The computation burden is reduced by the feasible examination from bottom to top and the goal optimization from top to bottom. An optimal method based on the back-tracing algorithm is presented for calculating the entire system restoration plan. The simulation results of the Shandong Power System in China demonstrate that the proposed decision-making structure fits to large-scale power system hierarchical dispatching and satisfies the coordination of various dispatching demand in different levels.