【摘要】 针对锂离子储能电站簇内电芯老化程度及其一致性难以准确评估的问题，提出基于特征数据信息熵的储能电站锂离子电池健康状态评估与预测方法。该方法将传统属性数据进行优化预处理以形成特征数据集，提出将信息熵概念移植到储能电站特定运行片段数据来展开分析，依据计算特征数据熵值大小情况来反映特征数据的有序程度，实现对簇内电芯老化程度及其一致性的分析判断，同时利用神经网络对熵值进行预测来对储能电站健康状态进行短期预测。最后通过储能电站实际运行数据与20S1P电池仿真模型验证基于特征数据信息熵值法对储能电站健康状态评估与预测的可行性与有效性，并在100 kW/200 kWh储能系统平台进行实际工程应用。更多还原

【Abstract】 In response to the challenge of estimating the aging degree and consistency of cells accurately in lithium-ion battery energy storage power station, this paper proposes a method for evaluating and predicting the health status of these cells based on the information entropy of characteristic data. This method involves optimizing and preprocessing traditional attribute data to form a characteristic data set. And it applies the concept of information entropy to analyze specific operation segment data of energy storage power stations innovatively. By calculating the entropy value of the characteristic data, the level of orderliness of the data can be determined, enabling analysis and assessment of the aging degree and consistency of cells within the cluster. Additionally, a neural network is utilized to predict the entropy value for health status short-term forecasting of the energy storage power station. The feasibility and effectiveness of this method, based on characteristic data information entropy for evaluating and predicting the health status of cells, are validated through simulation models of 20S1P cells and actual operation data from energy storage power plants. Furthermore, the method is applied in the actual engineering project involving a 100 kW/200 kWh energy storage system.更多还原