Electrochemical Battery State Estimation Under Parameter Uncertainty Caused by Aging Using Expansion Measurements

Accurate tracking of the internal electrochemical states of lithium-ion battery during cycling enables advanced battery management systems to operate the battery safely and maintain high performance while minimizing battery degradation. To this end, techniques based on voltage measurement have shown promise for estimating the lithium surface concentration of active material particles, which is an important state for avoiding aging mechanisms such as lithium plating. However, methods relying on voltage often lead to large estimation errors when the model parameters change during aging. In this paper, we utilize the in-situ measurement of the battery expansion to augment the voltage and develop an observer to estimate the lithium surface concentration distribution in each electrode particle. We demonstrate that the addition of the expansion signal enables us to correct the negative electrode concentration states in addition to the positive electrode. As a result, compared to a voltage only observer, the proposed observer can successfully recover the surface concentration when the electrodes' stoichiometric window changes, which is a common occurrence under aging by loss of lithium inventory. With a 5% shift in the electrodes' stoichiometric window, the results indicate a reduction in state estimation error for the negative electrode surface concentration. Under this simulated aged condition, the voltage based observer had 9.3% error as compared to the proposed voltage and expansion observer which had 0.1% error in negative electrode surface concentration.

I. Introduction

Lithium-ion batteries are ubiquitous in our portable computing devices and are playing a major role in the future of transportation with the transition to electric vehicles. To maintain a balance between power/energy demands and cost it is important to have an advanced battery management system that operates the battery safely, close to its limits, while minimizing the degradation. Accurate models and state estimation techniques are required to achieve this performance. The battery models can be classified as Equivalent circuit models (ECMs) and electrochemical models. ECMs are widely used in battery management system of electric vehicles because of their computational efficiency and state estimation using ECMs has been widely investigated [1], [2].