Information Technology Reference
In-Depth Information
«effector»
Move 1 step
«event»
Perceive
«event»
Move
«event»
Start
«sensor»
Perceive
«ev
ent»
Plan
«effector»
Plan
«event»
Pick-Up
«effector»
Pick-Up
«event»
Put-Down
«effector»
Put-Down
«event»
Communication
«sensor»
Receive
Knowledge
Figure A.3:
Sensor-effector-chains of a worker-agent.
of
Sensor-Effector-chains
. As can be seen, the sensor-effector-chains
combine the sensor and effector actions into a meaningful sequence.
Initially, each worker-agent receives a
Start
-event at the beginning
of a simulation. This event triggers the main loop of a worker-agent
consisting of the Perceive-sensor and the Plan-effector. This main
loop is executed by a worker-agent until the end of a simulation
no matter what happens. The Receive-sensor on the other hand
is only triggered by a Communication-event and stores incoming
communication messages into the agents' knowledge base.
The autonomous behavior of a worker-agent is defined completely
within the Plan-Action and quite simple: As long as no command is
given by the warehouse manager, worker-agents stay at their current
location and do nothing. As soon as the Receive-sensor of a worker-
agent is triggered, the worker-agent receives and stores the incoming
communication message into its internal knowledge base. Currently,
the possible set of communication messages is restricted to Transport-
Commands. Each Transport-Command consists of a point of origin
and a destination. First, a worker-agent has to move towards the
point of origin. Once it arrives it picks up the article which is located
there and transports it to the destination where it puts the article
down.
Figure A.4 shows the state chart of a worker-agent: As long as a
worker-agent has not received any command, it is in
Idle
-state. After
