Instrumentation & Measurement Magazine 25-7 - 18

[15]. Future EV fleets could also use constant connectivity for
on-demand, strategic charging instead of fast charging.
Improved BFG Algorithms: The research community is actively
studying and improving approaches to estimate BFG outputs.
As suggested in Fig. 4, more accurate BFG algorithms can help
to improve the RUL. Also, accurate SOC and TTS information
is crucial for the success of optimal charging algorithms.
Validation of BFG Algorithms
In addition to the need for improved BFG algorithms, there
are numerous challenges in evaluating a BFG. For example,
consider the estimated SOC values reported by a BFG; there
are no objective ways to find out the true SOC, even through
destructive means. In addition, the capacity (and hence the
SOC) might change with other environmental conditions as
well, making validation more challenging. The following
three metrics for BFG validation were presented and analyzed
in [17].
Coulomb Counting Metric (CC-metric): The SOC estimate by
the BFG is compared against the true SOC computed through
the Coulomb counting approach. Typical Coulomb counting
approach suffers from wrong initialization and incorrect battery
capacity assumption [4]. Here, the BFG validation profile,
shown in Fig. 5, is designed in a way that the battery capacity
can be retroactively computed. Similarly, the BFG validation
profile allows the Coulomb counting approach to have accurate
initial SOC information. The BFG is started at random
points in the BFG validation profile, denying both the initial
SOC and the opportunity to compute capacity.
Open Circuit Voltage Metric (OCV-metric): Here, the true SOC
is computed by resting the battery and performing an OCV
lookup [10]. The BFG validation profile is designed to have rest
periods to compute the OCV metric; the profile shown in Fig.
5 has two such points to compute the OCV metric. The BFG is
initiated at random points prior to the rest periods. For rigorous
evaluation, an OCV characterization should precede or
follow the BFG validation profile. This will allow to compute
the SOC more accurately at the rest point.
Time-to-voltage Metric (TTV-metric): The TTV metric computes
the error in the predicted time for the BFG to reach a
certain target voltage. Unlike the CC and OCV metrics, the
TTV metric is not a direct measure of SOC estimation error.
Rather, the TTV metric is a measure of prediction by the
BFG. Consider the remaining mileage reported by an EV. This
prediction can be easily verified by continuing to drive the
vehicle, preferable on a chassis dynamometer [18]. The TTV
metric is computed during a constant current as shown in the
BFG validation profile in Fig. 5.
In [17], all three BFG evaluation metrics were normalized to
be within 0% and 100%. For rigorous validation, these metrics
were computed at various temperatures spanning the entire
operational temperature region of a battery. For additional
rigor, the metrics must be computed for differently aged battery
samples.
Fig. 6 shows an experimental
hardware in the
loop setup used in for BFG
evaluation. The battery
is connected to a Kikusui
programmable load
that provided (pre-programmed)
load current.
The National Instrument
NI USB 6281 data acquisition
system recorded
the voltage across the battery
terminals and the load
current at 10 Hz sampling
rate. The temperature- controlled
bath with ethylene
glycol kept the battery at
a specified temperature
during the entire BFG validation
profile. Collected
data is used to compute the
BFG validation metrics at
different temperatures; a
detailed analysis was presented
in [17].
Fig. 5. Example of a BFG validation load profile. The CC metric is computed throughout the entire time; the OCV-metric at
the end of two rest periods; and the TTV-metric is computed during the last five hours (from [17], ©2015 Elsevier).
18
IEEE Instrumentation & Measurement Magazine
One of the limitations
in the BFG evaluation
October 2022

Instrumentation & Measurement Magazine 25-7

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 25-7