计及主配约束的分布式光伏承载力评估方法

A Distributed Photovoltaic Carrying Capacity Assessment Method for Calculating and Main Distribution Constraints

摘要: 文章旨在建立计及主配电网协同约束的光伏承载力评估方法，实现评估精度与效率的双重提升。通过构建含主配协同约束的评估模型，以最大光伏承载力为目标，将主网调峰缺口增量、净负荷最大波动率与配网电压、线损约束协同嵌入；针对传统全局寻优中海量时序潮流计算导致耗时过长这一核心痛点，提出了一种遗传算法(GA)与BP神经网络相结合的优化算法，通过BP神经网络构建“光伏装机容量–多维度约束指标”高精度非线性映射模型，通过模型预测与真实潮流计算验证，从而替代大量物理仿真。33节点算例验证表明，这种混合优化方法有效解决了优化过程中潮流计算耗时过长的问题，实现了配网–主网协同约束下光伏装机容量的高效、准确优化。

Abstract: The purpose of this paper is to establish a PV carrying capacity assessment method that takes into account the synergistic constraints of the main and distribution networks, and to realize the double improvement of assessment accuracy and efficiency. By constructing an assessment model with main and distribution synergistic constraints, with the maximum PV carrying capacity as the goal, the main grid peak gap increment, the maximum fluctuation rate of the net load and the distribution network voltage and line loss constraints are embedded synergistically; for the core pain point of the traditional global optimization in which the calculation of the massive time-sequence currents leads to the long time-consuming calculation, this paper proposes an optimization algorithm combining genetic algorithm (GA) and BP neural network, and constructs a high-precision nonlinear mapping model of “photovoltaic installed capacity-multidimensional constraint index” through BP neural network, and verifies it through model prediction and real current calculation, thus replacing a large number of physical simulations. 33-node case validation shows that this hybrid optimization method effectively solves the time-consuming and lengthy problem of the current calculation in the optimization process, enabling efficient and accurate optimization of PV installed capacity under coordinated transmission-distribution network constraints.