典型调峰/调频工况下储能电池组荷电状态估计
State of Charge Estimation of Energy Storage Battery Pack under Typical Peak/Frequency Modulation Conditions
朱沐雨 1马宏忠 1郭鹏宇 2宣文婧1
作者信息
- 1. 河海大学电气与动力工程学院,江苏南京 211100
- 2. 国网江苏省电力有限公司,江苏南京 210024
- 折叠
摘要
针对储能电池组在电网典型储能工况下荷电状态(state of charge,SOC)估算精度较低的问题,提出一种基于核主成分分析(kernel principal component analysis,KPCA)-鹈鹕优化(pelican optimization algorithm,POA)-双向门控循环单元(bidirectional gated recurrent unit,BiGRU)的SOC估计模型.通过设计调峰/调频工况下电池组充放电实验,从数据中提取表征SOC变化的融合特征作为模型输入;分别构建不同工况下BiGRU网络,并利用POA对其超参数进行优化,提高模型性能;进一步在混合工况下验证模型的有效性.结果表明,所建模型有着更好的SOC估计效果和更强的鲁棒性,能够提高复杂储能工况下储能电池组SOC估计精度.
Abstract
To address the issue of low estimation accuracy of the state of charge(SOC)for an energy storage battery pack under typical energy storage conditions of a power grid,this paper proposes a new SOC estimation model based on kernel principal component analysis(KPCA),pelican optimization algorithm(POA),and bidirectional gated recurrent unit(BiGRU).By designing the charge and discharge experiment of a battery pack under the condition of peak/frequency modulation,the paper extracts the fusion features of SOC change from the data as the model input.BiGRU networks are constructed under different working conditions,and POA is utilized to optimize its hyperparameters to improve the model's performance.The effectiveness of the model is further verified under mixed conditions.The results show that the proposed model has better SOC estimation performance and stronger robustness,which can improve the SOC estimation accuracy of energy storage battery packs under complex energy storage conditions.
关键词
储能电池组/荷电状态估计/调峰调频/鹈鹕优化/双向门控循环单元Key words
energy storage battery pack/state of charge estimation/peak and frequency modulation/pelican optimization/bidirectional gated recurrent unit引用本文复制引用
基金项目
国家自然科学基金(51577050)
国家电网江苏省电力公司科技项目(J2022158)
出版年
2024
国家科技期刊平台
NSTL
万方数据