The portable devices for battery health and charge monitoring are always
a challenge for engineers. This work is coin for development of
micro-controller based smart instruments to indicate status of battery.
It aims to develop an artificial intelligence based battery monitoring
instrument which indicates the charge status of the battery. The
instrument uses a Neuro-Fuzzy approach coupled with on line sensing of
parameters to map the battery behaviour. The battery behaviour shows
non-linear characteristics due to complex physical and chemical process
and is hard to predict. This behaviour is understood by modelling the
physical parameters. The soft computing techniques (artificial neural
network and fuzzy logic) have been used to model the battery behaviour
using easily measurable five parameters viz. voltage, current, internal
resistance, temperature and consumption time. The developed instrument
predicts battery status on line and indicates the battery status in
terms of state of charge that results in timely detection of faulty
state of the battery which is very essential for reliability and safety
of all instruments.