基于卡尔曼滤波算法的刀片电池温度估计

Blade Battery Temperature Prediction Based on Kalman Filter Algorithm

摘要: 大尺寸电池在运行过程中，温度不均匀性加剧，影响电池和电池组的性能、寿命和安全。然而，传感器和测试方法的局限性使得实时获取大尺寸电池的温度分布变得极其困难。为了快速预测大尺寸刀片电池的温度分布，本文提出了一种基于卡尔曼滤波算法的大尺寸电池温度估计方法。首先，在分析传热和产热机理的基础上，利用热平衡法建立了电池本体区域温度分布的差分方程。其次，利用卡尔曼滤波引入实测的温度形成闭环的校正机制，解决建模的误差和初始化的误差，该方法可以在短时间内获得电池五个节点区域的温度，实时预测大尺寸电池的温度分布演变。最后，通过对不同工况的测试，验证了所提出的方法用于温度分布预测的精度和效率。

Abstract: During the operation of large-size batteries, temperature unevenness increases, affecting the per-formance, life and safety of batteries and battery packs. However, the limitations of sensors and test methods make it extremely difficult to obtain the temperature distribution of large-sized batteries in real time. In order to predict the temperature distribution of large blade batteries quickly, a method based on Kalman filter algorithm was proposed. Firstly, on the basis of the analysis of the mechanism of heat transfer and heat production, the difference equation of the temperature dis-tribution in the region of the battery body is established by the thermal equilibrium method. Sec-ondly, the Kalman filter is used to introduce a closed-loop correction mechanism based on the measured temperature, which solves the modeling error and initialization error. This method can obtain the temperature of the five node areas of the battery in a short time and predict the temper-ature distribution evolution of large-sized batteries in real time. Finally, the accuracy and efficiency of the proposed method for temperature distribution prediction are verified by testing under dif-ferent working conditions.