| 引用本文: | 王阳,刘希喆.基于GRU-MPC的光储充电站日前-日内两阶段优化控制[J].电力自动化设备,2022,42(10): |
| WANG Yang,LIU Xizhe.Day-ahead and intra-day two-stage optimal control of photovoltaic-energy storage charging station based on GRU-MPC[J].Electric Power Automation Equipment,2022,42(10): |
|
| 摘要: |
| 电动汽车负荷与光伏出力的随机性给光储充电站的经济运行带来严峻考验。为了更高效地利用储能系统以及提高控制策略的鲁棒性，建立光储充电站日前-日内两阶段优化模型。在日前阶段，结合变分模态分解与门控循环单元神经网络预测一天48个时刻的充电负荷，并建立以日充电成本最小为目标的优化模型；在日内阶段，以日前调度计划与日内实际运行结果的偏差最小为目标，采用模型预测控制来实现滚动优化，为了增强储能控制策略对不确定源荷的跟踪能力，在日内关键时间点结合超短期源荷预测的结果对日前计划进行更新，得到基于阶段性最佳参考轨迹的实时调度。以实际算例进行仿真计算，比较不同控制策略对充电成本的影响，结果表明所提两阶段优化控制策略可以节省更多的充电成本，有更高的经济价值。 |
| 关键词: 充电站 日前优化 日内控制 变分模态分解 门控循环单元 模型预测控制 |
| DOI：10.16081/j.epae.202208036 |
| 分类号:U469.72 |
| 基金项目:国家自然科学基金委员会-国家电网公司智能电网联合基金资助项目(U2066212) |
|
| Day-ahead and intra-day two-stage optimal control of photovoltaic-energy storage charging station based on GRU-MPC |
|
WANG Yang, LIU Xizhe
|
|
School of Electric Power Engineering, South China University of Technology, Guangzhou 510641, China
|
| Abstract: |
| The randomness of electric vehicle load and photovoltaic output brings severe test to the economic operation of photovoltaic-energy storage charging station. In order to use the energy storage system more efficiently and improve the robustness of control strategy, a day-ahead and intra-day two-stage optimization model for photovoltaic-energy storage charging station is built. In the day-ahead stage, the variational mode decomposition and gate recurrent unit neural network are combined to predict the charging load at 48 moments of a day, and an optimization model with the minimum daily charging cost as its object is established. In the intra-day stage, the model predictive control is adopted to achieve rolling optimization with the minimum deviation between the day-ahead scheduling plan and the intra-day practical operation results as its object, in order to enhance the tracking ability of energy storage control strategy to uncertain sources and loads, the day-ahead plan is updated combining with the results of ultra short-term source and load prediction at the key intra-day moments, and the real-time scheduling based on staged optimal reference trajectory is obtained. The simulation and calculation are carried out with a practical example, the effect of different control strategies on the charging cost are compared, and results show that the proposed two-stage optimal control strategy can save more charging costs and has higher economic value. |
| Key words: charging station day-ahead optimization intra-day control VMD GRU MPC |
