Abstract: Dynamic reactive power optimization is an important measure to ensure power system operation economically and securely. On the basis of conventional dynamic reactive power optimization model, capacitor of the same node is equivalent to a variable, capacitor constraint instead of conventional single capacitor constraint, and completely equivalent action limit values and switching methods are given. With this model’s nonlinearity, massiveness, continuity and discreteness, ant colony algorithm used in discrete and nonlinear optimization is fully studied, and a hybrid algorithm which combines the interior-point method and ant colony algorithm is proposed. Test results on IEEE 14-bus systems demonstrate the efficiency and feasibility of this improved model and algorithm.

Key words: power system, dynamic reactive power optimization, ant colony algorithm, interior-point method

摘要： 动态无功优化是提高电力系统运行经济性和安全性的重要措施。在动态无功优化常规模型的基础上，将同一节点的若干组电容器等效为1个集中变量，用对集中电容器组的约束代替常规的单组电容器约束，给出与之完全等效的电容器组动作次数约束值和投切方法；针对动态无功优化模型的非线性、大规模性、控制变量的离散性和连续性共存的难题，对用于离散非线性优化的蚁群算法作了充分研究，提出了内点法与蚁群算法相结合的动态无功优化混合算法。IEEE14测试系统的仿真结果验证了改进模型及算法的有效性和可行性。

关键词: 电力系统, 动态无功优化, 蚁群算法, 内点法