基于ADMM的微电网群分布式无功电压优化控制方法研究

With the promotion of the electric power exchange marketization, geographically adjacent microgrids are communicating with each other through communication infrastructures for exchanging information about the states, control and trade, and are being interconnected to each other for power energy support, which drives those microgrids are being established as a cyber-physical microgrid cluster. This project will investigate distributed optimization algorithms for reactive power and voltage control methods of microgrid clusters with high penetration of distributed generation through controlling the inside microgrids or equipment of cluster level. Distributed optimization and control models are constructed considering multi-constrains and influence caused by cyber and physical sub-systems, and combining the analysis on operating characteristics of different microgrids. An optimization algorithm is proposed for reactive power based on the alternating direction method of multipliers (ADMM) and key factors affecting the convergence of the algorithm are investigated. Moreover, a distributed control method to coordinate the inside microgrids and equipment of cluster level is addressed under the cloud-event-triggering scheme to keep voltage stabilization and quality. To carry out the research on distributed optimization algorithms and coordinated control theory for microgrid clusters has great significance both in theory sense and engineering practice. And it also has practical significance to promote the large-scale application of distributed renewable resources under the national strategic planning of energy interconnection.

电力交易市场化的推进使地理上邻近的微网通过通信通道进行状态、控制、交易等信息的互动，通过互联能量通道进行电力互补支持，形成信息物理融合的微电网群。本项目面向分布式电源高渗透的微电网群，开展以其中的微网及群级设备为调控对象的分布式无功优化与电压协调控制方法研究。分析信息物理融合对微网无功电压的影响机制，考虑通信、电力、天气及环境等多种约束，并结合对不同微网运行特性的分析，构建分布式无功优化和电压协调控制模型；提出基于交替方向乘子法（ADMM）的分布式无功优化算法，探讨影响算法收敛性的关键因素；研究云事件触发机制下的分布式电压协调控制策略，实现各微网和群级设备在群控层面的协调运行，维持电压稳定，保证电压质量。针对微电网群开展分布式优化算法和协调控制理论的研究，具有重要的理论意义和应用价值；对于在能源互联的国家战略规划下，提高分布式可再生能源的大规模应用，推动经济与社会可持续发展具有现实意义。

随着科技的进步，以可再生能源为主的分布式发电单元在电网中的渗透率越来越高。以分布式电源及关键负荷为核心组件的微电网/微电网群/主动配电网通过信息网络传输交易、运行状态、控制指令等信息，通过物理互联交换功率。本项目面向分布式发电高渗透的信息物理融合微电网/微电网群/主动配电网，开展了以抑制分布式电源的随机扰动、提高可再生能源消纳能力的运行控制技术及优化调度方法的研究。考虑可再生能源发电的随机分布特性、负荷扰动的随机性、信息物理融合等因素，建立了DoS攻击下事件触发弹性负荷频率控制模型；分析得到深度学习算法在实现负荷频率控制时决策能力、环境适应度以及算法的收敛性定性结论，提出弹性触发机制及控制增益联合设计的方法及分布式能源并网逆变器改进虚拟同步控制技术，应用无需迭代的快速搜索与发现密度峰值聚类算法并结合同步回代消除法实现主动配电网电压分区协调控制问题。考虑发电成本、环境费用、并网收益、需求侧响应等因素，建立了不同目标组合约束下微电网优化调度模型，分析了粒子群算法、遗传算法、生物地理学优化算法等智能算法的全局搜索能力和迭代运行原理，提出加快智能算法寻优速度的思路及避免陷入局部最优的算法改进方法。