Reactive power sources have the potential for active distribution network (ADN) loss reduction and voltage profile improvement. However, the effective operation of reactive power sources depends upon their proper placement and sizing. To address this issue, this paper presents an optimal reactive power source allocation strategy in ADN. The salient feature of this strategy is that, by incorporating the ADN control scheme into the allocation process, it achieves the optimal allocation in a dynamic context. Specifically, smart transformer is considered as a particular realization of reactive power source, and the allocation problem is formulated as linear programming, which aims to minimize the investment cost of smart transformers. Meanwhile, the ADN control scheme is designed via model predictive control (MPC), with objectives of voltage regulation and loss reduction. Subsequently, the MPC formulation is embedded into the allocation problem, thus forming an integrated optimization problem, which links the control domain with the planning domain. By using the generalized Benders decomposition, this integrated problem is decomposed into a master problem and a set of subproblems, which can be solved alternately and iteratively. Simulation results verify the effectiveness of the proposed strategy.

中文翻译：

---

有源配电网中无功功率的最佳布置和尺寸：一种模型预测控制方法

无功电源具有减少有源配电网（ADN）损耗和改善电压分布的潜力。但是，无功功率源的有效运行取决于其正确放置和大小。为了解决这个问题，本文提出了一种在ADN中的最优无功电源分配策略。该策略的显着特征是，通过将ADN控制方案纳入分配过程，可以在动态上下文中实现最佳分配。具体来说，智能变压器被认为是无功功率的一种特定实现，分配问题被表述为线性规划，其目的是使智能变压器的投资成本最小化。同时，通过模型预测控制（MPC）设计ADN控制方案，以电压调节和降低损耗为目标。随后，将MPC公式嵌入到分配问题中，从而形成一个集成的优化问题，该问题将控制域与计划域链接在一起。通过使用广义的Benders分解，该集成问题被分解为一个主问题和一组子问题，这些子问题可以交替迭代地解决。仿真结果验证了所提策略的有效性。可以交替迭代地解决。仿真结果验证了所提策略的有效性。可以交替迭代地解决。仿真结果验证了所提策略的有效性。