Estimating Heat Generation Rate of Cylindrical Lithium-Ion Batteries by a Transfer Function

Measuring the heat generation rate of batteries is essential for their thermal management. To acquire the heat generation rate, the calorimeters are costly, while model-based methods are relatively low-cost but suffer from demanding data and/or compromised accuracy. To circumvent the dilemma, we develop a novel method with better accuracy to estimate the heat generation rate using only a single-point temperature by solving the inverse problem of the heat-generation-rate-temperature transfer function. Validated across different battery formats (18650 and 3270 liFePO 4 /graphite batteries), heat generation rates (from 0.5 C to 5 C), and ambient environments (oven, room, and silicone oil), we prove that this method accurately estimates the heat generation rate with either surface or internal temperatures measured by optical fiber sensors, offering ~56% less relative root mean squared error (RRMSE) than the conventional thermal network model. We also demonstrate another application of this approach: Using surface temperatures to derive the heat generation rate and then estimate internal temperatures, showing RRMSE below 3.2% at various discharge rates (0.5C-2C). This work not only facilitates the computation of the heat generation rate of batteries but also advances their thermal/thermodynamic studies together with battery sensors.