基于矩阵自适应校正的动态降维方法用于高风电渗透率下大规模电力系统电压相关暂态安全约束最优潮流

Transient security-constrained optimal power flow (TSCOPF) is an important class of problems for system operation. Several challenges arise when dealing with bulk power grids, including the large size and complex transient voltage behaviors. This paper aims to address such hurdles by proposing a dynamic dimensionality reduction matrix adaptive correction (DDR-MAC) algorithm, which can effectively evaluate proper Volt/Var levels to guarantee secure system operation. First, this paper performs dimensionality reduction processing at the bus and device levels to obtain a low-dimensional model with dominant modes, which solves the problems of high-order and large computational volumes of differential equations. Moreover, a dimensionality reduction error assessment model is established to ensure reduced-order accuracy. Then, the reduced-order TSCOPF model is equivalently decomposed into a mixed-integer linear optimization model and a combined coefficient correction model for system dynamic constraints and steady-state nonlinear constraints. Furthermore, a secant/tangent sensitivity adaptive correction method is presented to achieve fast computation. The DDR-MAC approach is verified across differently scaled IEEE test systems and the Nordic test system and can improve computational efficiency by 49.07% while offering higher accuracy than traditional computation methods.