作者：杨帆¹, 茅丰², 曹岑³

作者单位：1. 南通职业大学汽车与交通工程学院，江苏 南通 226007;
2. 上海应用技术大学电气与电子工程学院，上海 201418;
3. 国网南通供电公司，江苏 南通 226006

关键词：粒子滤波;锂动力电池;散热特性;生热机理;正则化

摘要：

由于当前锂动力电池散热特性研究方法受电池负载条件与使用环境的制约，导致测试精度低，采集结果不理想，该文提出基于粒子滤波改进算法的锂动力电池散热特性测试方法。基于锂动力电池的热生成机理，采用流体动力学方法构建锂动力电池散热系统传热数学模型。以此模型为状态方程对粒子滤波算法改进，采用自回归滑动平均模型确定锂动力电池散热的多步骤输出序列，并以输出值作为观测值建立锂动力电池散热状态空间模型。利用正则化粒子滤波算法进行重要性采样与重采样等过程，并迭代更新锂动力电池散热状态。实验结果表明：所提方法能准确获得锂动力电池使用过程中的温度变化与不同风道平均空气流速，不同测试阶段获得散热特性测试结果的相对误差分别为9%、4%和1%，均低于对比方法，锂动力电池平均使用寿命提高8.13%，说明该方法具有较高测试精度与推广价值。

Research on the test of heat dissipation characteristics of lithium battery based on improved particle filter algorithm
YANG Fan¹, MAO Feng², CAO Cen³
1. School of Automobile and Traffic Engineering, Nantong Vocational University, Nantong 226007,China;
2. School of Electrical and Electronic Engingeering Shanghai Institute of Technology, Shanghai 201418,China;
3. State Grid Nantong Power Supply Company, Nantong 226006,China
Abstract: Because of the limitation of the load condition and using environment, the measuring precision is low and the collecting result is not ideal. A testing method based on particle filter is put forward. Based on the thermal generation mechanism of lithium battery, the heat transfer mathematical model of heat dissipation system of lithium battery was established by hydrodynamic method. Using this model as the state equation to improve the particle filter algorithm, the multi-step output sequence of heat dissipation of lithium battery is determined by the autoregressive sliding average model. The process of importance sampling and resampling is carried out by using regularized particle filter algorithm, and the heat dissipation state of lithium battery is updated iteratively. The experimental results show that the temperature change and the average air velocity of different air ducts can be obtained accurately by the proposed method. The relative errors of the test results are 9%, 4% and 1%, respectively, which are lower than those of the comparison method. The average service life of the lithium battery is increased by 8.13%, which shows that the method has higher test accuracy and popularization value.
Keywords: particle filter;lithium battery;heat dissipation characteristics;heat generation mechanism;regularization
2020, 46(7):102-107  收稿日期: 2020-02-25;收到修改稿日期: 2020-04-08
基金项目: 上海联盟计划“农村电商智能物流配送系统”（LM201845）
作者简介: 杨帆(1980-),女,江苏苏州市人,讲师,硕士,主要研究方向为汽车电气及其智能控制、监控及自动控制技术等
参考文献
[1] 姜贵文, 黄菊花, 刘明春, 等. 锂离子动力电池模块设计及散热特性[J]. 电源技术, 2018, 42(6): 782-784, 817
[2] SHULIN L, NAXIN C, YAN L, et al. Modeling and state of charge estimation of lithium-ion battery based on theory of fractional order for electric vehicle[J]. Transactions of China Electrotechnical Society, 2017, 32(4): 189-195
[3] 许爽, 苏玉, 王伟. 电动汽车锂离子电池组内散热特性的数值模拟[J]. 科学技术与工程, 2018, 18(16): 55-61
[4] 张凝, 徐皑冬, 王锴, 等. 基于粒子滤波算法的锂离子电池剩余寿命预测方法研究[J]. 高技术通讯, 2017, 27(8): 699-707
[5] 秦帅星, 郭超超. 锂离子电池热解气体释放特性研究[J]. 消防科学与技术, 2017, 36(11): 1592-1595
[6] 路昭, 余小玲, 张立玉, 等. 锂电池组高温节点空气冷却方案的数值模拟[J]. 西安交通大学学报, 2018, 52(7): 25-31, 138
[7] JIANG Z, ZHOU W, LI H, et al. A new kind of accurate calibration method for robotic kinematic parameters based on the extended Kalman and particle filter algorithm[J]. IEEE Transactions on Industrial Electronics, 2017, 65(4): 3337-3345
[8] 江健. 改进的粒子滤波算法在船用组合导航中的应用[J]. 上海海事大学学报, 2018, 39(2): 17-21
[9] 谢金法, 谢宁, 李博超. 纯电动汽车锂离子动力电池热特性研究[J]. 电源技术, 2019, 43(6): 1001-1004
[10] 应炳松, 贺元骅. 不同加热功率触发锂离子电池热失控特性研究[J]. 工业安全与环保, 2018, 44(6): 14-16, 57
[11] 郑玺, 李新国. 基于OpenCV的组合优化多目标检测追踪算法[J]. 计算机应用, 2017, 37(S2): 112-114, 145
[12] 张威虎, 郭明香, 贺元恺, 等. 一种改进的蝴蝶算法优化粒子滤波算法[J]. 西安科技大学学报, 2019, 39(1): 119-123
[13] 罗景文, 秦世引. 基于Dirichlet过程非参贝叶斯学习的高斯箱粒子滤波快速SLAM算法[J]. 机器人, 2019, 41(5): 660-675
[14] LIPU M S H, HANNAN M A, HUSSAIN A, et al. State of charge estimation for lithium-ion battery using recurrent NARX neural network model based lighting search algorithm[J]. IEEE Access, 2018, 6: 28150-28161
[15] 胡坚强, 舒志兵. 基于均值偏移与粒子滤波融合的目标跟踪算法研究[J]. 电子器件, 2019, 42(3): 706-711