零碳园区电-氢混合储能系统多目标优化配置

随着氢气生产和储存技术的快速发展,开发氢气储能系统(hydrogen energy storage systems, HESSs)将给能源和电力系统结构带来根本性变化。HESSs和电池储能系统(battery energy storage systems, BESSs)相结合进行协调优化可以解决多种能源供需之间的不平衡,并提高能源效率。为确保BESSs和HESSs规划的有效性,以最小全生命周期成本(life cycle cost, LCC)、系统网损、联络线交换功率偏差、负荷波动以及电压波动为目标,采用带精英策略的非支配排序遗传算法(non-dominated sorting genetic algorithm-II, NSGA2)求解储能系统(energy storage systems, ESSs)选址定容规划方案的Pareto非支配解集。并利用基于熵权法(entropy weight method, EWM)的灰靶决策在Pareto非支配解集中选取最佳折中解。另外,通过模糊核C-均值(fuzzy kernel C-means, FKCM)聚类算法获取源荷典型运行场景集,并基于扩展的IEEE-33节点系统进行仿真分析。仿真结果表明:NSGA2算法不仅实现了电-氢混合储能系统LCC最小,且其电压质量、功率稳定性、网损与负荷波动也显著优于对比算法。

Multi-objective Optimal Configuration of Electricity-Hydrogen Hybrid Energy Storage System in Zero-Carbon Park

With the rapid development of hydrogen production and storage technology, the development of hydrogen energy storage systems (HESSs) will bring fundamental changes to energy and power system structure. The coordinated optimization of HESS and battery energy storage system (BESS) can solve the imbalance between supply and demand of various energy sources and improve energy efficiency. In order to ensure the effectiveness of BESS and HESS planning, minimum life cycle cost (LCC), system network loss, switching power deviation, load fluctuation, and voltage fluctuation are chosen as the fitness function in this paper. Meanwhile, a non-dominated sorting genetic algorithm-II (NSGA2) with elite strategy is used to solve energy storage system (ESS) Pareto non-dominated solution set of site-constant volume planning scheme. The grey target decision based on entropy weight method (EWM) is used to select the best compromise solution in Pareto non-dominated solution set. Additionally, typical operation scenarios of source load are obtained by fuzzy kernel C-means (FKCM) clustering algorithm, and the simulation analysis is carried out on the basis of the extended IEEE 33-node system. Simulation results show that NSGA2 algorithm not only achieves the minimum LCC of the electricity-hydrogen hybrid energy storage system, but also improves voltage quality, power stability, network loss and load fluctuation compared to that of other algorithms.