风光储柴独立混合微电网能量多目标优化及电压稳定性研究

The stand-alone hybrid microgrid with renewable energy generation has a great significance to the normal power supply of the remote mountainous areas or the independent power supply of the special occasions. This project will make further research on energy multi-objective optimization and voltage stability study for stand-alone hybrid microgrids. Firstly, taking the reactive compensation factors into account, we construct the model of the multi-objective optimization allocation of distributed generation, and put forward a Markov-based electromagnetism-like mechanism algorithm to improve the solving efficiency of algorithm. For the uncertain units in the hybrid microgrid, the multi-objective dynamic scheduling optimization strategy that has the immunity of the data uncertainty is proposed, a two-stage multi-objective robust optimization unit commitment model is also built, as well as the model is solved by using the generalized Benders decomposition principle and the PSO algorithm to get the robust optimal solution which has a closed form through the rolling optimization. Synthesizing the upper dynamic scheduling strategy of the micro power, a multi-objective optimization distributed stability control strategy based on the energy storage devices and DSTATCOM is also proposed, to realize the whole optimization and the coordination control of the system. Finally, we use the branch model of simplified electric network to build the model of the hybrid microgrid, and study the dynamic bifurcation characteristic of voltage instability of the system based on the bifurcation theory to get the online stability criterion of the system. The research results of the project can provide a theoretical basis for the independent、economic、steady and reliable operation of the hybrid microgrid system.

含可再生能源发电的独立混合微电网对偏远山区正常供电或特殊场合独立供电具有重要意义。本项目对独立运行混合微网能量多目标优化理论与系统稳定性分析进行深入研究。构建考虑无功补偿因素的分布式电源多目标优化配置模型，并提出基于Markov的仿电磁学算法来提高算法求解效率；针对混合微网中不确定单元，提出对数据不确定性具有免疫力的多目标优化动态调度策略，构建两阶段多目标鲁棒优化微电源机组组合模型，基于广义Benders 分解原理和PSO算法对模型进行求解，通过滚动优化获得具有闭环形式的鲁棒最优解；集成上层微电源动态调度策略，构建基于储能装置及DSTATCOM的多目标优化稳定控制策略，实现系统整体优化与协调稳定控制；采用简化电网络支路模型对混合微网进行建模，基于分岔理论研究系统电压失稳的动态分岔特性，获得系统运行的在线稳定性判据。项目研究成果可为混合微网系统独立、经济、稳定、可靠地运行提供理论依据。

针对以风光发电为主的微电网在能量优化管理调度及稳定性和可靠性研究中还存在诸多不足，本课题对微电网分布式电源多目标优化配置、鲁棒优化动态调度、电压稳定性理论分析及分布式电源控制策略等方面展开了相关研究。. 研究内容及成果包括：（1）针对独立运行微电网，以经济性、环保性和能源利用效率综合最优为目标，引入熵理论确定多目标函数的权重系数，构建了电源容量优化配置模型，将Markov 模型嵌入改进的仿电磁学算法中以提高模型求解效率；针对并网运行微电网，考虑系统自平衡度和冗余度等约束，构建了并网型微网优化配置模型，采用改进生物地理学优化算法进行求解。（2）针对风光发电的不确定性，构建了多场景机会约束动态调度模型；构造以预测值为中心的不确定性集，引入鲁棒优化理论，以灰熵关联度为评价标准选择多目标最优解，构建了基于储能运行状态的含不确定性预测调节策略的鲁棒优化动态调度模型；采用滚动时域优化并结合短时预测的方法来改进动态调度方法，从而提高了微网调度的鲁棒性；针对风光出力及负荷的预测偏差，基于序列运算理论优化了微电网正负旋转备用容量。（3）将分岔理论应用于孤立微电网分析其静态电压稳定性，以鞍结分岔作为电压稳定临界点，研究负荷和无功补偿装置对鞍结分岔点的影响；提出一种基于曲率半径及两分搜索法的改进延拓法来追踪系统的平衡解流，求出系统的鞍结和霍普夫分岔点，该方法使平衡解流形在弯曲程度高区域失真较小，在平坦区域追踪速度较快。（4）针对孤立直流微网，研究了基于储能的稳压控制方法。针对交流微网，研究了光伏发电、风力发电并网控制方法；研究了基于无功电流一致控制的自适应虚拟阻抗下垂控制，以及基于虚拟同步发电机的主从控制；探讨了基于分组牵制一致性算法的微网群分布式协调控制；研究的分布式电源接入微网的电压调节系统及稳压装置获得发明专利授权。. 已发表或录用学术论文31篇（8篇被EI收录），发明专利授权3项，培养硕士生12名、博士生2名。