| Abstrak/Abstract |
Energy problems have become a recent issue that
has occurred in recent years. This problem occurs due to the
number of vehicles continues to increase vehicle motor every
year. Electric vehicles have a primary energy source comes from
batteries. To support battery function optimally and prevent the
battery from short-circuit, good Battery Management System
(BMS) is needed for fault detection. Therefore, appropriate
battery model and fault detection algorithms are needed to
perform fault detection based on the characteristics of the fault
type. In this study, a connection fault (short circuit) detection
algorithm based on Luenberger Obsever has been developed, in
which the input signal (current input) that affects the Observer
system fault model and the thevenin model are used to describe
the characteristic of the battery current. The connection fault
scenario based on change in value of current input on terminal
output voltage. Simulation was carried out with the fault occured
at t = 4000 s to 6000 s and there is one time fault. On the other
hand, there ware multi time fault occured at t = 1000 s to 2000
s, t = 3000 s to 3400 s, and t =4000 s to 6000 s. Form the
simulations that the fault can be detected by sensor readings
at the output terminal voltage at any time and in several time
intervals. Connection fault as current input error is represented
by residual signal obtained by Luenberger Observer algorithm |
