Increasing integration of photovoltaic (PV) system in electric grids cause congestion during peak power feed-in. Battery storage in PV systems increases self-consumption, for consumer’s benefit. However with conventional maximising self consumption (MSC) control for battery scheduling, the issue of grid congestion is not addressed. The batteries tend to be fully charged early in the day and peak power is still fed-in to grid. This also increases battery degradation due to increased dwell time at high state of charge (SOC) levels. To address this issue, this work uses a model predictive control (MPC) for scheduling in PV system with battery storage to achieve multiple objectives of minimising battery degradation, grid congestion, while maximising self consumption. In order to demonstrate the improvement, this work compares the performances of MPC and MSC schemes when used in battery scheduling. The improvement is quantified through performance indices like self consumption ratio, peak power reduction and battery capacity fade for one-year operation. An analysis on computation burden and maximum deterioration in MPC performance under prediction error is also carried out. It is concluded that, compared to MSC, MPC achieves similar self consumption in PV systems while also reducing grid congestion and battery degradation.

中文翻译：

---

基于光伏的微电网中使用模型预测控制的电网拥塞缓解和电池退化最小化

越来越多的光伏 (PV) 系统在电网中的集成会导致高峰供电期间的拥堵。光伏系统中的电池存储增加了自我消耗，为消费者带来利益。然而，对于电池调度的常规自耗最大化 (MSC) 控制，电网拥塞问题并未得到解决。电池往往在当天早些时候充满电，峰值功率仍会馈入电网。由于在高荷电状态 (SOC) 水平下的停留时间增加，这也会增加电池退化。为了解决这个问题，这项工作使用模型预测控制 (MPC) 在具有电池存储的 PV 系统中进行调度，以实现多个目标，即最大限度地减少电池退化、电网拥塞，同时最大限度地提高自耗。为了证明改进，这项工作比较了 MPC 和 MSC 方案在用于电池调度时的性能。改善通过一年运行的自耗率、峰值功率降低和电池容量衰减等性能指标量化。还对预测误差下的 MPC 性能的计算负担和最大恶化进行了分析。得出的结论是，与 MSC 相比，MPC 在光伏系统中实现了类似的自耗，同时还减少了电网拥塞和电池退化。