Real-time Lithium-ion (Li-ion) battery state estimation based on electro-thermal model

点击翻译按钮获取中文摘要

Lithium-ion (Li-ion) batteries are essential components for storing electricity and delivering electric power on demand in different applications, such as electronics, electric vehicles, and power grid storage. The reliability of these storage systems is strongly dependent on the safety and robustness of the batteries. It rests on the use of efficient battery management systems (BMSs) that are usually employed to monitor, control, and manage the thermal performance of the batteries, as well as check their safety and integrity. The BMSs perform these activities based on the assessment of some battery internal state conditions, like State-Of-Charge (SOC) and State-Of-Health (SOH), that are not measurable but estimated based on measurable signals (like voltage, current, and temperature) as the battery is charged or discharged. Nowadays, almost all commercial battery state estimators rely on a battery empirical model and Kalman Filter (KF). The empirical-based models are fast, but they are not accurate enough as the battery ages, especially in harsh ambient conditions wherein the physical parameters representing its behavior change rapidly. Furthermore, such models cannot estimate the maximum battery temperature, which is an important factor for battery safety. The proposed research program aims at developing diagnostic and control tools for the sustainable management of energy storage systems. The short-term objective is to develop an improved Li-ion battery state estimator based on a more representative physic-based model. The internal battery states like SOC or SOH will be estimated by using a reduced model coupled to an extended Kalman Filter (EKF), due to the non-linearity of the governing equations. In this program, the battery electrical signals and the surface temperature are measured to update the battery internal states in the EKF and for estimating internal temperature distribution. The methodology relies on a combination of mathematical modeling, battery testing, and material characterisation. The work is divided into 6 steps: 1- development of a preliminary physic-based model, 2- design, construction and use of a battery testing experimental bench, 3- model optimisation and estimation of its thermo-physical parameters, 4- battery post-mortems and electrode characterisation, 5- development of aging equations to consider the impact of degradation mechanisms, 6- coupling the new model with EKF-based estimators to assess battery's SOC and SOH. After 5 years, 7 HQPs will have been trained in a domain where Canada has important needs. It will bring the development of new knowledge and technologies for the blossoming Canadian energy storage industry. The new tools developed to improve the energy management of Li-ion batteries can also be applied to boost the economic development of clean energy industries in Canada, ranging from electric automotive engines to electricity storage needed in the production and distribution of renewable energies