Diagnosis of Battery Condition Using Ultrasonic Measurements

Summary:

In lithium-ion batteries (LIBs), mechanical alteration of the composites can cause a safety-critical state during battery operation. These alterations can be caused by deformation, inhomogeneous straining, or lithium plating just to name a few. Besides a safety risk, such alterations can lead to a massive loss of capacity, accelerate the aging processes and vastly limit the fast charging capability.
Common diagnostic algorithms, based on voltage measurement, temperature measurement and coulomb counting, can only determine the state of charge (SoC), state of health (SoH), and lithium plating at specific laboratory circumstances. Complex mechanical processes in the battery can not be tracked, which provide valuable information about the battery state.
This new measurement technique aims to increase the accuracy of the battery management system (BMS) state estimation procedures, with regards to SoC/SoH determination and detection of safety-critical states by sensing propagation of ultrasonic guided waves. Capturing the mechanical properties of the battery can further enable early detection of mechanical alterations and thus prevent total battery damage.
In recent years, a large number of signal analysis methods have been published. To react with very high sensitivity to slight changes in the ultrasonic signal, very sophisticated algorithms have to be used. On the other hand, concerning the automotive industry, it is necessary to run a lean algorithm on a small, low-cost embedded system in real-time. To fulfill both requirements, an optimized version of the so-called matching pursuit algorithm was developed and implemented in this work.
The matching pursuit decomposes the ultrasound signal into a linear combination of signals of reduced complexity, which are then analyzed. A very low computational complexity could be reached so that the processing of the ultrasound data is possible in real-time on an embedded device.
By using the matching pursuit, strong correlations regarding SoC and expansion were found in various battery formats.