微电网集成性可调节鲁棒并行优化运行研究

In order to give full play to comprehensive efficacy of microgrids in the aspects of improving energy efficiency, reducing air pollution emissions and improving power quality, this project studies on the microgrid optimal operation strategy from the viewpoint of integrating global optimal dispatch with distributed power control, and the strong supports for the integrated strategy implementation are mainly provided from the following three aspects: (1) Taking impacts of uncertainties like renewable energy power into full account, the research extends the modeling theory of microgrid optimal dispatch to the category of robust multi-objective optimization with adjustable uncertainty interval, and achieves flexible control to robustness of the optimal solutions by introducing the concept of uncertainty budget, to compensate for deficiencies of conventional robust optimization which is conservative and difficult to solve. (2) In order to efficiently solve the complex optimization model above, the research studies on L-J potential individual diversity automatic recovery mechanism which is based on molecular dynamics, and construct multi-core-population parallel multi-objective evolution algorithm, to break through the bottleneck of searching depth and speed within existing optimization algorithms. (3) To satisfy the real-time on-line control requirements of the microgrid integrated optimal operation system, the research introduces multi-core cluster parallel computing technology into the process of microgrid coordination optimal control and robust multi-objective optimization computation, so as to drastically promote control efficiency and computation speed. The research can provide new thought for the microgrid optimal operation modeling and calculation, which has important theoretical value and practical significance.

为充分发挥微网在提高能源效益、减少污染排放和改善电能质量等诸多方面的综合功效，从全局性优化调度和分布式电能控制相集成的角度研究微网优化运行策略，并主要从以下三方面为该集成性策略实施提供有力支持：(1)充分考虑可再生能源发电等不确定性因素的影响，将微网优化调度建模拓展到不确定区间可调节鲁棒多目标优化范畴，通过引入不确定性预算概念实现对优化解鲁棒性的灵活调控，以弥补常规鲁棒优化偏于保守以及难以求解的不足；(2)为高效求解上述复杂模型，研究基于分子动力学的L-J势个体多样性自动恢复机制，并构造多核种群并行多目标进化算法，以突破现有优化算法在寻优深度和速度上的瓶颈；(3)为满足微网集成性优化运行系统在线控制实时性的要求，研究将多核集群并行运算技术引入到微网协调优化控制和多目标鲁棒优化计算过程，以大幅提升控制效率和计算速度。课题研究可为微网优化运行建模及计算提供新的思想，具有理论价值和现实意义。

为充分发挥微网在提高新能源效益、减少污染排放和改善电能质量方面的综合功效，从微网多目标优化调度与电能控制系统相结合的角度研究微网集成性优化运行策略具有必要性和重要性。若是将微网多目标优化调度系统与电能控制系统相结合，系统计算复杂度势必大为提高，如何满足系统运行在线控制的实时性要求则是首要面临的问题。此外，还需要探索一种更为有效的鲁棒多目标优化建模及计算方法，以便能得到对微网运行中不确定性因素的影响具有较强鲁棒性的优化解，这无疑会进一步增加优化计算的复杂度，又给上述实时性问题的解决带来了新的挑战。.本项目旨在从微网鲁棒多目标优化调度与电能控制系统相结合的角度，探索一种微网集成性优化运行策略，并能对可再生能源发电等不确定性因素的影响具有灵活可调的鲁棒性。研究内容主要包括微网集成性鲁棒并行优化运行策略研究、微网可调节鲁棒多目标优化调度模型研究、微网并行协调优化控制体系研究、并行多目标优化算法研究、多核集群并行运算环境研究等方面。.项目充分考虑可再生能源发电等不确定性因素的影响，将微网优化调度建模拓展到不确定区间可调节鲁棒多目标优化范畴，通过引入不确定性预算概念实现对优化解鲁棒性的灵活调控，弥补了常规鲁棒优化偏于保守以及难以求解的不足；为高效求解上述复杂模型，基于分子动力学的L-J势个体多样性自动恢复机制构造了多核种群并行多目标分子微分进化算法，突破了现有优化算法在寻优深度和速度上的瓶颈；为满足微网集成性优化运行系统在线控制实时性的要求，将多核集群并行运算技术引入到微网协调优化控制和多目标鲁棒优化计算过程，大幅提升了控制效率和计算速度。.本项目研究最终有效解决了微网优化运行所面临的集成性、鲁棒性和实时性的关键问题，为微网系统运行建模及优化控制、多目标优化算法发展提供了新的概念和思想，具有理论意义和学术价值；对于实现节能减排、优化电能结构和推动新能源发电产业规模化具有现实意义及应用价值。