Measurement Coulomb counting

Comparison between state of charge determination methods for lithium-ion batteries based on the measurement of impedance, voltage and coulomb counting... 

Many battery monitoring systems rely on voltage, current, and temperature measurements advanced devices also include coulomb counting. Sensors provide SOC information and identify irregularities, but capacity estimations are beyond reach for most battery management systems (BMSs). 

A key element in management of hthium batteries is the accurate and reliable determination of battery state-of-charge (SOC), whether or not the "history" of the battery is known from date of manufacture. Rehable SOC measurements are critical for many devices. Well-known interrogation techniques used to measure battery SOC include coulomb counting [11], voltage recovery [12], and electrochemical impedance spectroscopy (EIS) [13]. Often, these approaches yield data that is highly nonlinear, complex, or intractable, especially for rechargeable battery systems. 

Primary Li/SOa cells are important for applications such as the SINCGARS radio. We have developed two fuzzy logic-based approaches to determining the SOC of these cells - one suitable for an internal meter based on Coulomb counting and another suitable for an external meter based on ceU impedance measurements. 

Molecules that are ionized by electron impact in the ion source are accelerated, sent through a conventional 90° magnetic sector analyzer, postaccelerated by a few thousand volts, and arrive at the electron multiplier detector. The output of the electron multiplier detector consists of pulses of about lO- coulomb per ion. The pulses are amplified and sent through a gated amplifier and an electronic switch which is synchronized with the beam chopper so that one of the ion counters records ions only when the beam chopper is open, the other only when the beam chopper is closed. The difference between the two ion counts represents the ion intensity contributed by the molecular beam, while the square root of the sum of the two ion counts is approximately equal to the standard deviation of the measurement and serves as a useful indicator of the quality of the data being obtained. 

Two topical issues may be mentioned. The first is the definition of the potentials that are measured by different techniques, say by AFM, electrokinetically and externally imposed, and their relationships [11], The second is of a more theoretical nature and concerns modeling of the nondiffuse part of the double layer. The classical approach is through Stem theory [2], which in most cases is adequate, although it requires two additional parameters. A more recent development is in terms of ion correlations, essentially an advanced statistical theory whereby all coulombic ion-ion and ion-surface interaction pairs are counted and statistically summed [2]. This is a step forward over the smeared-out models of Gouy and Stern. The issue here is that cases must be found where deviations from Gouy theory cannot be interpreted on the basis of the Stem model... 

---