基于改进的NSGA-II重力储能重物块分级优化策略

张强; 李建文; 马明晗; 管春宝; 马明; 唐骞骏; 郦昶成; 徐紫嫣

doi:10.19799/j.cnki.2095-4239.2025.1015

您当前的位置：

首页 >

文章列表页 >

基于改进的NSGA-II重力储能重物块分级优化策略

储能系统与工程 | 更新时间：2026-04-29

- 基于改进的NSGA-II重力储能重物块分级优化策略
- Weight-block gradation optimization for gravity energy storage based on non-dominated sorting genetic algorithm II
- 储能科学与技术 2026年15卷第4期页码：1331-1342
- 作者机构：
  
  燕赵电力实验室(华北电力大学)，河北保定 071003
- 作者简介：
  
  张强（2002—），男，硕士研究生，研究方向为重力储能技术及其应用，E-mail：15548201922@163.com
  李建文（1983—），女，副教授，研究方向为重力储能技术及其应用，E-mail：ljw_ncepu@163.com；
- 基金信息：
  
  中央高校基本科研业务费专项资金资助项目(2024MS111)
- DOI：10.19799/j.cnki.2095-4239.2025.1015
  中图分类号： TM 921.2
- 收稿：2025-11-12，
  
  修回：2026-01-28，
  
  纸质出版：2026-04-28
- 稿件说明：
移动端阅览
张强, 李建文, 马明晗, 等. 基于改进的NSGA-II重力储能重物块分级优化策略[J]. 储能科学与技术, 2026, 15(4): 1331-1342.

ZHANG Qiang, LI Jianwen, MA Minghan, et al. Weight-block gradation optimization for gravity energy storage based on non-dominated sorting genetic algorithm II[J]. Energy Storage Science and Technology, 2026, 15(4): 1331-1342.
张强, 李建文, 马明晗, 等. 基于改进的NSGA-II重力储能重物块分级优化策略[J]. 储能科学与技术, 2026, 15(4): 1331-1342. DOI： 10.19799/j.cnki.2095-4239.2025.1015.

ZHANG Qiang, LI Jianwen, MA Minghan, et al. Weight-block gradation optimization for gravity energy storage based on non-dominated sorting genetic algorithm II[J]. Energy Storage Science and Technology, 2026, 15(4): 1331-1342. DOI： 10.19799/j.cnki.2095-4239.2025.1015.

摘要

兆瓦级重力储能系统中采用电励磁电机直接并网，可有效利用电机固有特性实现暂态电压与频率的支撑，但单一质量等级重物块的储放势能无法平滑连续地调节新能源并网系统的波动性功率，因此对重物块质量分级优化是进行小步长功率调节的重要工程方法。本研究针对该工程实际问题，提出一种改进的非支配排序遗传算法II(NSGA-II)的重物块分级策略：将相互冲突的重物块分级数量最少与平均功率补偿误差最小两个指标作为优化目标，考虑重物块质量边界约束以及单时段重物块操作次数约束等工程实际条件，建立重物块质量分级的多目标优化模型，结合工程约束引入贪心局部搜索策略，确保优化结果满足实际要求，完成重力储能分级的工程实际问题求解。为提高解集可靠性，执行10次独立运行优化，采用超体积、拥挤度、理想点距离等多维评价指标结合TOPSIS方法综合评价各次运行结果，筛选出最优Pareto前沿，并以平均功率补偿误差不超过5%作为工程决策依据选取出分级数最少的最优方案，针对春、夏、秋、冬四季典型日的风、光、负荷功率波动场景，实现百兆瓦时重力储能最优重物块分级组合，最后通过蒙特卡洛方法生成8个随机扰动场景进行仿真验证本分级方案的鲁棒性和有效性。

Abstract

In megawatt-level gravity energy storage systems

utilizing electrically excited motors for direct grid connection effectively leverages inherent motor characteristics to provide transient voltage and frequency support. However

using a single mass-grade weight block to store and release gravitational potential energy cannot provide smooth

continuous regulation of power fluctuations in a renewable-energy grid-connected system. Therefore

optimizing weight-block mass grading is a critical engineering approach toward fine-step power regulation. To address this practical issue

this study proposes an improved non-dominated sorting genetic algorithm II-based weight-block grading strategy. The optimization targets are two conflicting objectives: minimizing the number of weight-block grades and minimizing the average power-compensation error. Considering practical engineering constraints

such as mass boundaries of weight blocks and operational frequencies per block within a single period

a multi-objective optimization model is established for weight-block mass grading. A greedy local search strategy is incorporated to handle these engineering constraints

thereby ensuring that the optimization results meet practical requirements and effectively address the grading problem of gravity energy storage. To enhance the reliability of the solution set

10 independent optimization runs are executed. Subsequently

multi-dimensional evaluation metrics

including hypervolume

crowding distance

and ideal point distance

are combined with the technique for order of preference by similarity to ideal solution (TOPSIS) to comprehensively evaluate each run and identify the optimal Pareto front. Using an average power-compensation error not exceeding 5% as the engineering decision criterion

the scheme with the minimum number of grades is selected as the optimal solution. For typical everyday wind

solar

and load-power fluctuations across spring

summer

autumn

and winter

an optimal weight-block grading combination is achieved for a 100 MWh gravity energy storage system. Finally

the robustness and effectiveness of the proposed grading scheme are validated via Monte Carlo simulations using eight randomly generated disturbance scenarios.

关键词

Keywords

references

闫文举, 王洋, 孙芯竹, 等. 基于直线电机的废弃矿井重力储能系统研究进展与关键技术[J]. 储能科学与技术, 2025, 14(1): 255-268.

YAN W J, WANG Y, SUN X Z, et al. Research progress and key technology of abandoned mine gravity energy storage system based on linear motor[J]. Energy Storage Science and Technology, 2025, 14(1): 255-268.

喻恒凝, 姚良忠, 程帆, 等. 重力储能在新型电力系统中应用: 前景及挑战[J]. 中国电机工程学报, 2025, 45(18): 7177-7192. DOI:10.13334/j.0258-8013.pcsee.240834.

YU H N, YAO L Z, CHENG F, et al. Prospects and challenges of gravity energy storage applications in new type power system[J]. Proceedings of the CSEE, 2025, 45(18): 7177-7192. DOI:10.13334/j.0258-8013.pcsee.240834.

李震, 陈巨龙, 李文林, 等. 提升斜坡式重力储能AGC性能的混合储能优化运行方法[J]. 储能科学与技术, 2024, 13(8): 2761-2771. DOI:10.19799/j.cnki.2095-4239.2024.0211.

LI Z, CHEN J L, LI W L, et al. Optimized operation of hybrid energy storage to enhance the performance of AGC with sloped gravity storage[J]. Energy Storage Science and Technology, 2024, 13(8): 2761-2771. DOI:10.19799/j.cnki.2095-4239.2024.0211.

李震, 王斌, 牟雪鹏, 等. 基于瞬时功率镜像补偿的斜坡式重力储能系统功率平滑控制策略[J]. 电机与控制应用, 2024, 51(5): 12-20. DOI:10.12177/emca.2024.026.

LI Z, WANG B, MU X P, et al. Power smoothing control strategy for slope gravity energy storage system based on instantaneous power mirror compensation[J]. Electric Machines & Control Application, 2024, 51(5): 12-20. DOI:10.12177/emca.2024.026.

DEB K, PRATAP A, AGARWAL S, et al. A fast and elitist multiobjective genetic algorithm: NSGA-II[J]. IEEE Transactions on Evolutionary Computation, 2002, 6(2): 182-197. DOI:10.1109/4235.996017.

YANG X S. Multiobjective firefly algorithm for continuous optimization[J]. Engineering with Computers, 2013, 29(2): 175-184. DOI:10.1007/s00366-012-0254-1.

邵振, 邹晓松, 袁旭峰, 等. 基于改进多目标粒子群优化算法的配电网削峰填谷优化[J]. 科学技术与工程, 2020, 20(10): 3984-3989. DOI:10.3969/j.issn.1671-1815.2020.10.028.

SHAO Z, ZOU X S, YUAN X F, et al. Optimization of peak load shifting in distribution network based on improved MOPSO algorithm[J]. Science Technology and Engineering, 2020, 20(10): 3984-3989. DOI:10.3969/j.issn.1671-1815.2020.10.028.

王先齐, 吕智林, 汤泽琦. 基于分时电价机制的并网型微网多目标动态优化调度[J]. 电力系统保护与控制, 2017, 45(4): 9-18. DOI:10.7667/PSPC160250.

WANG X Q, LYU Z L, TANG Z Q. Multiobjective dynamic optimal dispatching of grid-connected microgrid based on TOU power price mechanism[J]. Power System Protection and Control, 2017, 45(4): 9-18. DOI:10.7667/PSPC160250.

李永毅, 王子晗, 张磊, 等. 风-光-氢-燃气轮机一体化氢电耦合系统容量配置优化[J]. 中国电机工程学报, 2025, 45(2): 489-501. DOI:10.13334/j.0258-8013.pcsee.232133.

LI Y Y, WANG Z H, ZHANG L, et al. Capacity allocation optimization of integrated hydrogen-electric coupling system of wind-solar-hydrogen-gas trubine[J]. Proceedings of the CSEE, 2025, 45(2): 489-501. DOI:10.13334/j.0258-8013.pcsee.232133.

李琦, 韩运滨, 白章, 等. 离网型风光发电联合制氢系统的协调运行策略与容量配置优化[J]. 中国电机工程学报, 2024, 44(20): 8136-8145, I0018.

LI Q, HAN Y B, BAI Z, et al. Coordination operation strategy and capacity optimization of off-grid wind-solar hybrid hydrogen production system[J]. Proceedings of the CSEE, 2024, 44(20): 8136-8145, I0018.

李翠萍, 司文博, 李军徽, 等. 基于集合经验模态分解和多目标遗传算法的火-多储系统调频功率双层优化[J]. 电工技术学报, 2024, 39(7): 2017-2032. DOI:10.19595/j.cnki.1000-6753.tces.230186.

LI C P, SI W B, LI J H, et al. Two-layer optimization of frequency modulated power of thermal generation and multi-storage system based on ensemble empirical mode decomposition and multi-objective genetic algorithm[J]. Transactions of China Electrotechnical Society, 2024, 39(7): 2017-2032. DOI:10.19595/j.cnki.1000-6753.tces.230186.

史沁鹏,曾小超,李建文,等.MW级固体重力储能系统设计及多机协同运行策略[J/OL].南方电网技术,1-11[2025-06-24]. https://nfdwjs.csg.cn/gateway-web/zh/debutDetail.html?serialNum=20240927004.

曾小超, 姜健宁, 李建文, 等. 百MWh级竖井式重力储能系统重物块储存与运输方案[J]. 储能科学与技术, 2025, 14(10): 3839-3847. DOI:10.19799/j.cnki.2095-4239.2025.0196.

ZENG X C, JIANG J N, LI J W, et al. Storage and transportation schemes for a one-hundred-megawatt-hour-class shaft-type gravity energy storage system[J]. Energy Storage Science and Technology, 2025, 14(10): 3839-3847. DOI:10.19799/j.cnki.2095-4239.2025.0196.

HWANG C L, YOON K. Multiple attribute decision making[M]. Berlin, Heidelberg: Springer, 1981 DOI:10.1007/978-3-642-4831 8-9.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

百MWh级竖井式重力储能系统重物块储存与运输方案

重力储能技术研究进展及展望

直线电机重力储能系统充放电策略优化

基于电动汽车无序充电行为的可再生能源系统综合评估及调控策略研究