Agents in Biology
203
possesses internal memory which would guide them between resource and
nest-site locations.
The ant agents exit a nest site when their hunger level drops below a
threshold value and they begin searching for food. Agents deposit pheromones
in the environment as they walk and sense the presence of pheromone nearby.
If a pheromone trail is present they follow the trail, but if multiple pheromones
trails are detected, the ant chooses whcih trail to follow based on concentrations found. Once a food source is located, the ant agent eats 0.02 units of
food and stops movement.
If there are no pheromones nearby to follow, then the ant agents move
randomly, but include a probabilistic turning matrix known as a ‘turning
kernel’. This ‘turning kernel’ is derived from an actual video of Pharaoh’s
ants, where digital analysis or video tracking was used [23].
The body length of a Pharaoh’s ant, 2-mm is represented as 2 units in the
model. This is depicted by a single step that an ant agent takes simulating a
movement of 2-mm. A realistic foraging space for a typical colony was identified as 25-cm to 150-cm. The models were simulated in an environment of
500-mm by 500-mm.
The environment contains a ‘nest’ agent placed at the centre of the environment and two food source agents. FLAME also supports dynamic creation
of agents allowing pheromone agents to be created when required by the ant
agents during the simulation, or when an ant agent takes a step.
Figure 7.12 depict the trials generated by the ant agents during the simulation when two food sources are available. The blue trail depicts the ant
paths and the yellow depicts the pheromone deposits along the paths.
Find Nest without Pheromone Smell Model
First version
Probability to leave active trail added 0.001
Mesude Bicak
270110
noPheromoneSmell.c
Ant
int antID
double antX
double antY
double foodLevel
int isFed
int isInNest
int antDirection
int state
int pheroFound
203
possesses internal memory which would guide them between resource and
nest-site locations.
The ant agents exit a nest site when their hunger level drops below a
threshold value and they begin searching for food. Agents deposit pheromones
in the environment as they walk and sense the presence of pheromone nearby.
If a pheromone trail is present they follow the trail, but if multiple pheromones
trails are detected, the ant chooses whcih trail to follow based on concentrations found. Once a food source is located, the ant agent eats 0.02 units of
food and stops movement.
If there are no pheromones nearby to follow, then the ant agents move
randomly, but include a probabilistic turning matrix known as a ‘turning
kernel’. This ‘turning kernel’ is derived from an actual video of Pharaoh’s
ants, where digital analysis or video tracking was used [23].
The body length of a Pharaoh’s ant, 2-mm is represented as 2 units in the
model. This is depicted by a single step that an ant agent takes simulating a
movement of 2-mm. A realistic foraging space for a typical colony was identified as 25-cm to 150-cm. The models were simulated in an environment of
500-mm by 500-mm.
The environment contains a ‘nest’ agent placed at the centre of the environment and two food source agents. FLAME also supports dynamic creation
of agents allowing pheromone agents to be created when required by the ant
agents during the simulation, or when an ant agent takes a step.
Figure 7.12 depict the trials generated by the ant agents during the simulation when two food sources are available. The blue trail depicts the ant
paths and the yellow depicts the pheromone deposits along the paths.