基于Barzilai-Borwein梯度法的多微电网系统递阶优化调度算法

针对由多个相互关联的微电网组成的多微电网系统，本文提出基于Barzilai-Borwein梯度法的多微电网系统递阶优化调度算法。该算法通过协调各子网多余或缺少的能量，并优化各子网发电设备以实现各微电网能量供需平衡，最小化多微电网系统整体运行成本。首先建立多微电网能量流动模型及其能量管理优化问题，进而综合预测控制的滚动优化思想和关联平衡原理，设计了多微电网系统的递阶优化调度算法。该算法将整体优化的复杂的高维整体优化问题分解为若干个相互独立的低维子问题，将原来集中的大规模计算分散到每个子系统中进行，降低对单个控制器的计算性能要求，并引入Barzilai-Borwein梯度法对子微电网之间协调，提升协调级迭代计算寻优过程的收敛速度，驱动各子网的独立优化解快速收敛至多微电网的整体优化解，以较低的计算量取得了良好的多微电网能量管理效果。最后通过数值仿真验证了本文所设计算法的有效性。

The hierarchical optimization algorithm of multi microgrid system based on Barzilai-Borwein gradient method

In this paper, a hierarchical optimization scheduling algorithm based on Barzilai-Borwein gradient method is proposed for the multi-microgrid system consisting of multiple interconnected microgrids. The algorithm fulfills the energy supply and demand balance of each microgrid by coordinating the excess or lack of energy and optimizing the power generation of each subnet, and makes the overall operating cost of the multi-microgrid system to be lowest. Firstly, the energy flow model of multi-microgrid and its energy management optimization problem are established. Then the rolling optimization idea and the principle of correlation balance are combined to design a hierarchical optimization scheduling algorithm. The algorithm decomposes the complex high-dimensional optimization problem into several independent low-dimensional sub-problems, and distributes the centralized large-scale calculations into each subsystem to reduce the computational burden for a single controller. The Barzilai-Borwein gradient method is introduced to coordinate the sub-microgrid, and speed up the convergence speed of the coordination process. The independent optimization solution of each subnet is driven to quickly converge to the overall optimization solution of the multi-microgrid. The proposed method achieves better energy management effect with lower computational burden. Finally, the effectiveness of the proposed algorithm is verified by numerical simulation.