Fuzzy model of foraging behavior in a bird flocking simulation

The folloeing work presents the approaches taken while extending a fuzzy model for a computer simulation of bird flocking to a fuzzy model, that allows the simulation of bird foraging, as well. In order to extend the original model it was necessary to extend the artificial world and the artificial bird. The artificial world was extended with the introduction of feeding areas--circular regions containing food. Similar to the artificial birds the feeding areas were implemented as animats, i.e. by means of a transition function. The first stage is responsible for the selection of important information about the birds that are currently within the feeding area. We extended the artificial birds with the introduction of notion of hunger. In order to do this two additional drives were added to the model's basic drives. The artificial bird's nest internal state was extended as well. Three properties wew added: hunger, height and mode. In order to support the attraction of the feeding areas a new percepcion function was added and the action selection was extended so as to take into account the influences of the newly introduced drives on bird's hunger, height, mode, which can be in one of four states: feednig, take-off, flying and landing. While the mode drive is responsible for the transition between these modes, the feeding drive influences hunger, height and speed and flight direction of the bird with respect to the current mode. With the extensions made we brought our model a step closer to a realistic simulation of bird behaviour.