大容量锂离子电池系统可信容量估计及其偏差机理研究

In order to solve some problems in high-efficiency operation and control of large-scale Li-ion battery energy storage system(BESS), such as the physical relationship between battery system performance parameters and its operating characteristics can not be illustrated accurately and the battery system capacity credit estimation is imprecision, the project carries out research on the capacity credit estimation and its deviation principle of the large-scale Li-ion battery system. The project will select LiFePO4 batteries as the research object. The cells performance parameters variation mechanisms and the model parameters simplified methods in multi-time scales application will be studied. The rules of battery system model parameters expansion will be established exactly by a screening process method and a parameters corrector. An accurate battery system model in multi-time scales will be built using equivalent circuit method by the applicant.The project will investigate the state of charge(SOC) and its deviation principle of the battery system with high dynamic and nonlinear characteristics when the prior noise statistics are unkown or inaccurate. An adaptive unscented Kalmal filter(UKF) algorithm with a suboptimal and recurred noise statistical estimator is presented to estimate the battery system capacity credit. A battery system and a test platform of its capacity credit estimation will be set up to validate the theories and methods presented in the project.The project will not only increase capacity credit estimation accuracy of the large-scale battery system, but provide theoretical basis for optimizing configuration and control of battery system，and facilitate the large-scale battery energy storage system efficient and stable operation and its application.

针对大容量（MW级）锂离子电池储能系统高效运行与控制中存在难以准确表征电池系统性能参数与其工作特性的物理关系、电池系统可信容量估计精度不高等问题，本项目开展大容量锂离子电池系统可信容量估计及其偏差机理研究。拟以磷酸铁锂电池为研究对象，探讨不同时间尺度下电池单体串并联时电池单体性能参数变化机理及其模型参数优化方法，运用筛选过程法及电池性能参数校正器来确立电池系统模型参数扩展规律，采用等效电路法建立多时间尺度大容量电池系统数学模型；研究噪声先验统计信息未知或不明确时高动态、非线性大容量电池系统荷电状态估计及其偏差机理，利用带次优递推噪声统计估计器的自适应无迹卡尔曼滤波法进行大容量电池系统可信容量估计；搭建电池系统及其可信容量估计平台，验证所提理论与算法。本项目的研究可提高大容量电池系统可信容量估计精度，为电池系统的优化配置与控制提供理论依据，将促进大容量电池储能系统稳定高效运行及其工程化应用。

针对大容量（MW级）锂离子电池储能系统高效运行与控制中存在难以准确表征电池系统性能参数与其工作特性的物理关系、电池系统可信容量估计精度不高等问题，本项目开展了大容量锂离子电池系统可信容量估计及其偏差机理研究。以磷酸铁锂电池为研究对象，通过准确表征大容量电池系统中电池性能参数与其工作特性的物理关系，采用等效电路法分别建立长时间尺度与短时间尺度的电池单体模型，结合筛选过程法与电池性能参数校正器建立了串联电池系统模型、并联电池系统模型及串并联型电池系统模型；研究了噪声先验统计信息不准确时大容量电池系统荷电状态估计及其偏差机理，提出了两种不同的基于噪声统计估计器的自适应无迹卡尔曼滤波法来进行大容量电池系统荷电状态估计；搭建了系统测试平台，验证了所提算法的准确性和有效性。项目研究成果提高了大容量电池系统建模及其可信容量估计精度，为电池系统的管理与控制提供理论与技术支撑，可促进大容量电池储能系统稳定高效运行及其工程化应用。