Ultra-Fast Temperature Estimation of Lithium-ion Batteries Through Impedance Measurements

Abstract

Lithium-ion (Li-ion) batteries have become increasingly important in the electrification of transportation and electrical energy production in recent years. Despite the growing popularity, Li-ion batteries face several challenges such as safety hazards and environmental issues from battery waste. Safe and long-lasting battery operation can be achieved by efficiently monitoring and regulating key state parameters such as the cell temperature. However, current temperature estimation methods rely on limited sensoring devices and complicated modeling techniques which can be difficult to implement at low cost. Studies have shown a good correlation between the battery impedance and temperature, suggesting that the former can be a suitable indicator for estimating the latter. The recently proposed discrete-interval binary sequence (DIBS) broadband method have been shown to overcome several disadvantages of traditional electrochemical impedance spectroscopy such as long measurement time and costly equipment. This paper demonstrates the use of the DIBS in developing a temperature estimation algorithm for a commercial Li-ion battery cell. Experimental results based on several commercial Li-ion battery cells are presented to demonstrate the effectiveness of the proposed method. An overall estimation accuracy of less than 2 °C absolute error was achieved by applying least square fitting and linear models, thus showing that temperature estimation through cell impedance measurement can be highly effective even when applying simple modeling techniques.