基于改进非线性预报-校正内点法的最优潮流

提出一种改进的非线性预报-校正内点法,该方法在校正阶段应用了多中心校正的超立体空间映射技术,经过合理配置一些关键参数改善了映射空间的结构,使算法在每次迭代时只需1次校正计算就能获得较大的迭代步长,从而快速收敛至最优解.通过IEEE 14,IEEE 30,IEEE 57,IEEE 118这4个测试系统的仿真计算表明,该算法收敛快,其迭代次数基本与每次迭代进行4次中心校正计算的多中心校正内点法相当,而且鲁棒性好,未出现数值稳定问题.

Improved Nonlinear Predictor-corrector Interior Point Method for Optimal Power Flow

An improved nonlinear predictor-corrector interior point method is proposed. A hypercube projection technology which has been applied in multiple-centrality-correction interior point method (MCC-IPM) is introduced to the correction stage of the proposed method. The structure of projection space is improved by setting some key parameters rationally to make the method get good iteration step lengths only by one correction calculation. So the method can converge to the optimal solution rapidly. The numerical performance of the proposed method is evaluated on four power systems ranging in size from 14 buses to 118 buses. Computational results demonstrate that this technique is fast, robust, and its iteration time is equal to the MCC-IPM with four centrality-correction for every iteration.