Abstract:
In the 21st century, there has been a significant increase in demand for electric vehicles
and energy storage technology. To meet this demand, various technologies have been
developed, including lithium-ion batteries, sodium-ion batteries, and vanadium flow batteries.
However, many high-energy batteries face challenges related to material longevity and safety.
As a result, it is crucial to incorporate a battery state determination algorithm into battery
management systems. There are several parameters that determine the battery's state, such as
its health, charge level, power status, function, and safety.
In this manuscript, a new algorithm is developed for assessment of the state of health of
the battery through various techniques. The secondary and primary batteries were used for
creating the state of health estimation model. As a secondary battery, a popular Li ion (NCM)
cylindrical battery referred to as NCR18650 (Panasonic was selected and as a primary battery
cylindrical lithium thionyl battery (Minamoto) was selected.
In this research work, the state of charge of a battery was determined based on the
thermodynamics parameters of the battery which is fitted through equation: SOC = 𝛼 + 𝛽 ∗
Δ𝐻 + 𝛾 ∗ Δ𝑆, where Δ𝐻 − enthalpy, Δ𝑆 – entropy, SOC – state of charge, and 𝛼, 𝛽, 𝛾 −are
coefficients which depend on state of health and chemistry of a battery [1]. on of the
dependence of coefficients to the state of health of the battery was done. One of the problems
during the thermodynamics parameters measurement includes its long measurement time.
Therefore, more fast measurement method of entropy and enthalpy was developed.
Additionally, the state of health of the secondary battery was analyzed by and
emerging technique of machine learning. A suitable algorithm and feature was chosen to
assess state of health of the battery based on charge profile data.