| Electrification has been the most viable way for achieving clean and efficient transportation as demonstrated in hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), and battery electric vehicles (BEVs). As the most prominent energy storage device for these advanced electrified vehicles, batteries have been intensively investigated and used in recent years. However, under specific operating conditions, such as fast acceleration and regenerative braking, high current magnitudes can be delivered from or to the battery cells. Consequently, the large amount of heat generation caused by over-charging, over-discharging, and over-temperature may bring potential permanent damages to the batteries. In this respect, accurate monitoring of key operational parameters, such as current, voltage, temperature, state of charge (SOC), and state of health (SOH), is fundamental to the design and control of an effective battery management system (BMS).;In this dissertation, existing SOC and SOH estimation techniques are analyzed first with emphasis on their merits and demerits as applied to automotive applications. Based on the concept of impulse response, a precise online SOC estimation method is proposed. Furthermore, two pattern recognition techniques for detection of SOC are presented and compared. Experimental validations have been conducted on two battery systems.;Continuous monitoring of temperature distribution throughout the battery is crucial to determination of the SOH for the battery system. This dissertation also explores different temperature distribution estimation methods, namely, using thermodynamic and heat transfer principles in the battery, based on which a three-dimensional (3D) electro-thermal model has been developed and validated through multiple simulations and experiments. However, due to the computational time and system requirements, this numerical model cannot be employed for real time monitoring of temperature distribution in an online thermal management application. Therefore, a temperature estimation model based on virtual thermal sensors (VTS) for the Li-ion battery is proposed next. This model, using a small number of physical sensors, is able to estimate temperature distribution throughout the battery in real time under unknown initial values and model uncertainty. Experimental results from a high energy Li-ion battery validate the effectiveness of the proposed VTS model under various charging and discharging conditions. In addition, a new temperature estimation method based on thermal impulse response is also developed which doesn't require in-depth knowledge of battery chemical and structural properties. This method is able to accurately predict the temperature variations on the surface of a Li-ion battery. |
