Alpha Wolf Hunt Behavior

Wolves hunt in packs of about 5 to 20 members and keep a certain hierarchy while eating a captured prey with the stronger alpha wolf eating first. This organization permits the wolves to hunt animals that are even larger than their own size.

The wolf pack hunting model described in this work includes the following important assumptions:

a)  Wolf teams are conformed by a group leader (alpha wolf) and at least one follower (beta wolf).

b)  Beta wolves group around the alpha wolf keeping a certain distance from the leader and among themselves.

c)  Wolves in the model receive only visual information from the environment, using this input to calculate their positions and distances. Any other type of communication between wolves is not allowed.

d)  Visual fields are limited to a single camera recognizing objects by their colors. If alpha wolf is outside beta wolf visual field, then beta wolves loses track of leader.

e)  Head direction is kept constant relative to body motions.

f)   Walking speeds are kept constant for all wolves at all times.

In the following sections we describe the formation, hunting and avoidance model in more detail.

Hunt Behavior

Alpha Wolf

The alpha wolf behavior is determined by three states: Wander, Stalk, Attack and Eat.

  • Wander. In this state the alpha wolf explores the environment looking for a prey to eat. When it detects one, the prey_visible condition is activated indicating a change to the Stalk state.
  • Stalk. In this state two transitions can happen: one towards the Wander state in case the prey is outside its range of vision caused by the prey_not_visible condition; the other one occurs when the prey is detected close, activated by the prey_near condition continuing to the Attack state.
  • Attack. In this state the wolf closes on the prey until catching it. When this happens the prey_catch conditionis activated and the wolf changes to the Eat state.
    If the wolf suddenly loses control of the prey it goes back to the Stalk state activated by the prey_not_catch condition.
  • Eat. In this state the wolf eats the prey. After that, the condition of prey not visible is activated indicating that the prey has been eaten. The wolf goes back to the Wander state.

Fig. 3. Consists of four states: Wander, Stalk, Attack and Eat. Each state is activated by the corresponding conditions.

Beta Wolves.

The beta wolf behaviour is described by five states, Wander, Formation, Stalk, Attack and Eat.

  • Wander. The beta wolf searches the environment looking for the group leader or prey. If the wolf finds the leader, leader_visible is activated continuing to the Formation state. If the wolf finds the prey, prey_visible is activated continuing to the Stalk state. The wolf movement is similar to the alpha wolf in this state.
  • Formation. As long as the beta wolf continues seeing the alpha wolf, it stays close to the leader. If visible contact is lost with the leader then the condition leader_not_visible is activated continuing to the Wander state. If the beta wolf detects a prey it moves to the Stalk state since the condition prey_visible is activated.
  • Stalk. In this state three transitions can happen: one towards the Formation state in case the prey is outside its range of vision activated by the prey_not_visible condition; another one occurs when it detects the prey near, this activating the prey_near condition, continuing to the Attack state; and, in case the beta wolf loses track of the alpha wolf, it executes the leader_not_visible condition is activated, continuing to the Wander state.
    The objective of this state is to approach the prey without separating much from the group leader.
  • Attack. The beta wolf considers itself sufficiently close to the prey although still taking into consideration the relative position to the leader.
    Yet, in this state the move to approach and catch the prey has greater priority than follow the group leader.
    Nevertheless, because the prey is in many cases faster the prey could escape. If this happens the prey_far condition is activated and the beta wolf returns to the Stalk state. If the wolves manage to catch the prey, the condition catch_prey is activated continuing to the Eat state.
  • Eat. This state is similar to the alpha wolf although alpha wolves eat first.

Fig. 4. Consists of five states: Wander, Formation, Stalk, Attack and Eat. Each state is activated by the corresponding conditions.

Avoid Behavior

Although wolves are by nature predators, they can also be predated upon, e.g., by humans and other wolves.

  • Hunt. This behavior encompasses all behaviors previously described in Figure 3 and 4 for alpha and beta wolves, respectively.
  • Avoid. At any time during hunting, the appearance of apredator will make the alpha and beta wolves run away in direction opposite to the predator.

Fig. 5. Avoid is activated by a predator_visible condition while in the Hunt state, and the other way with a predator_not_visible condition.

Lähde: A Biologically-Inspired Wolf Pack Multiple Robot Hunting Model

Kategoria(t): Riistanhoito Avainsana(t): , , , . Lisää kestolinkki kirjanmerkkeihisi.