Panasonic Eco Solutions Nordic AB
MEW01622
Rev: -
EBL128 Planning Instructions V2.0.x
94
13.4.2
Filtering algorithm
In order to secure a fast detection of real fire alarms and to reduce
nuisance (false) alarms to a minimum, a filtering algorithm is used.
The filtering algorithm uses the sensor values to calculate a decision
value depending on which alarm algorithm that is in use. The
decision value is zero from the beginning. Each time a new sensor
value is picked up (sampled) from an analog smoke detector 430x /
440x, this new sensor value is compared with the actual decision value
and the decision value will be adjusted or not adjusted as follows:
If the difference, between the new sensor value and the actual decision
value is < "
X
", the decision value is set equal to the new sensor value.
If the difference is > "
X
", the decision value is increased or reduced
by "X".
"
X
" = The Step Value. It is different depending on the sensitivity and
detection time, i.e. it is depending on the selected alarm algorithm, see
Figure 18.
The decision value can consequently not be increased / decreased with
a value exceeding the "X" value even if the sensor values are much
higher / lower.
Analog
detector
Normal detection time (15sec.)
Slow detection time (35sec.)
H-15
2.4%,
High
sensitivity
N-15
3.0%,
Normal
sensitivity
L-15
3.6%,
Low
sensitivity
H-35
2.4%,
High
sensitivity
N-35
3.0%,
Normal
sensitivity
L-35
3.6%,
Low
sensitivity
4300 / 4301
4400 / 4401
X=8
X=10
X=12
X=4
X=5
X=6
Figure 18. The six alarm algorithms for the detectors 430x and
440x in NORMAL mode. Default is alarm algorithm
N-15
, i.e.
normal detection time (15 sec.) and normal sensitivity (3%).
X=The step value.