Fuzzy logic system is the translation of the accident
detection/action system. It represents the formulas that
govern the system in a pure mathematical view or fuzzy
logic view. What makes Fuzzy logic an excellent tool is
the ability to make quick decision in uncertain conditions
that normally gives more than satisfactory results to the
problem, this applies precisely to accident detection in
Intelligent Traffic Light Systems. Accident detection is
usually sensed by the disturbance it causes to the
surrounding lanes, in other words if the accident didn’t
cause any disturbance in lane the Intelligent Traffic System
won’t bother reacting to it because it does not affect the
flow of the traffic which is the main concern of Traffic
Light system. Detecting the accident visually is currently
challenging and limited therefore cannot be used in the
system. Other arguments can be made regarding the
consequences of leaving the accident unresolved which
might cause troubles to the flow later on, this discussion
will be left out as the main focus here is resolving
accidents that affect the current traffic flow and not the
predicted one. The disturbance in traffic flow is the main
sign that an accident has occurred and needs reaction. This
simple conclusion requires inputs that are related to the
road, the vehicles and data that can only be concluded by
combining the two factors together i.e. cross ratio.
Additionally those inputs must be mapped to appropriate
values since they will not be treated as precise inputs but
rather estimated duo to the nature of road, so a membership
function will assign the inputs to the appropriate values.
After that, fuzzy Rules will take those inputs as facts then
depending on the mapped inputs and the rule an action will
be taken that will improve the traffic flow or at least
prevent it from getting worse. . As seen in fig. (3) fuzzy
logic system is composed of the following elements: