Impedance Based Temperature Estimation of Lithium Ion Cells Using Artificial Neural Networks

Tracking the cell temperature is critical for battery safety and durability. It not feasible to equip every with a sensor in large systems such as those electric vehicles. Apart from this, sensors are usually mounted on surface do detect core temperature, which can mean detecting an offset due gradient. Many sensorless methods require great computational effort solving partial differential equations or error-prone parameterization. This paper presents estimation method lithium ion cells using data electrochemical impedance spectroscopy combination artificial neural networks (ANNs). By training ANN of 28 estimating temperatures eight more same type, network (a simple feed forward only one hidden layer) was able achieve accuracy ΔT= 1 K (10 ∘C <T< 60 ∘C) low effort. The estimations were investigated different types at various states charge (SoCs) superimposed direct currents. Our easy use be completely automated, since there no significant monitoring temperature. In addition, prospect above mentioned approach estimate additional SoC state health (SoH) discussed.