【摘要】 提出了一种基于对偶松弛法的扩展凸规划和序列线性规划相结合的在线最优潮流方法。通过分析核心约束交换逻辑的物理内涵,使优化过程清晰可理解;通过一个独立的探测和处理不可行解的预过程,使核心优化计算摆脱了不可行问题;通过引入控制效能指标,使低效控制解的数目减少,增强了结果的可操作性;通过扩展凸规划方法,使各种可分目标得到有效处理,部分实现了目标函数和算法自身的分离。该方法可灵活应用于在线安全校正控制及在线优化调度。在几个测试系统及一个省级实际电网上进行了数值试验,结果表明该方法是快速可靠的。更多还原

【Abstract】 Based on the dual relaxation method, an extended convex programming and sequential linear programming optimal power flow approach is presented for online optimal dispatch. The optimization process is easy to be understood and the proposed approach is of the ability of detecting infeasible solutions in advance that makes the core optimization process to be disengaged from infeasibility problem. Key factors, which influence the effectiveness of control variables, are analyzed. Based on these factors, an index is defined to compress low efficiency controls, and to enhance the usability of optimal solutions. Extended convex programming method can easily handle all kinds of convex separable objectives and partially realize the separation of objectives and optimization process. This method can be flexibly applied to power system online corrective control and optimal dispatch. Numerical tests in several systems and an actual 121 buses system are given in the paper.更多还原