1 The Behavior of Schools Groups of fishes can behave almost like a - - PDF document

1

15/11 - 05 1 Emergent Systems, Jonny Pettersson, UmU

Last time

Self-Organization Autonomous Agents Real Ants Virtual Ants Ant Algorithms Assignment 2 Assignment 3

15/11 - 05 2 Emergent Systems, Jonny Pettersson, UmU

Outline for today

Schooling of fish Boids

15/11 - 05 3 Emergent Systems, Jonny Pettersson, UmU

Flocks, Herds, and Schools

”and the thousands of fishes moved as a

huge beast, piercing the water. They appeared united, inexorably bound to a common fate. How comes this unity?”

• Anonymous, 17th century

2

15/11 - 05 4 Emergent Systems, Jonny Pettersson, UmU

The Behavior of Schools

Groups of fishes can behave almost like a

single organism

Trafalgar effect

Rapid transfer of information Can execute swift, evasive maneuvers Reaction propagates many times faster than

the approach of the predator Ex: Flash expansion and the fountain

effect

Predators can also coordinate their

movements

Ex: Parabolic formation of Giant bluefin tuna

15/11 - 05 5 Emergent Systems, Jonny Pettersson, UmU

The Behavior of Schools

Individuals rarely collide, even in frenzy of

attack or escape

Shape of school is characteristic of

species, but flexible

Herring 3:3:1 (ratio of length: width: depth) Pollack 6:3:1 Cod 10:4:1 to 2:4:1

Arrangement within schools is also

characteristic of species

Depend also on the size and the speed of the

school

15/11 - 05 6 Emergent Systems, Jonny Pettersson, UmU

Adaptive Significance

Prey avoiding predation

Safety in numbers United erratic maneuvers Pattern of body coloration Group breaking behaviors Compact aggregation – predator risks injury by

attacking

3

15/11 - 05 7 Emergent Systems, Jonny Pettersson, UmU

Adaptive Significance

Better predation

Coordinated movements – tuna More efficient predation

• Killer whales encircle dolphins
• Take turns eating

Schooling may increase hydrodynamic

efficiency

Endurance may be increased up to 6 times

V-formation of geese

Range increase 70%

Lobster line up

Move 40% faster – decreased hydrodynamic

drag

15/11 - 05 8 Emergent Systems, Jonny Pettersson, UmU

Behavior of Individuals within the School

A balance between attraction and repulsion Sensory inputs

Vision – attraction and alignment The lateral line system – repulsion and speed

matching Weighting of information coming

simultaneously from several fishes

Most strongly influenced by nearest neighbors

15/11 - 05 9 Emergent Systems, Jonny Pettersson, UmU

Alternatives to Self-Organization

Templates

No evidence that water currents, light or

chemicals guide collective movement Leaders

No evidence for leaders Those in front changes Each adjusts to several neighbors

Blueprint or recipe

Implausible for coordination of large schools

4

15/11 - 05 10 Emergent Systems, Jonny Pettersson, UmU

Self-Organization Hypothesis

Simple rules generate schooling behavior

Positive feedback – brings individuals together

(attraction)

Negative feedback – but not to close (repulsion)

Only local information

Positions and headings of a few nearby fish No leader, no global plan

15/11 - 05 11 Emergent Systems, Jonny Pettersson, UmU

Huth and Wissel (1992) Model – Basic Assumptions

Each fish follows the same rules Each fish uses some form of weighted

average of the position and orientation of its nearest neighbors

Each fish responds to its neighbors in a

probabilistic manner

Imperfect information gathering Imperfect execution of actions

No external influences affecting the fish

No currents, obstacles...

15/11 - 05 12 Emergent Systems, Jonny Pettersson, UmU

Huth and Wissel (1992) Model – Behaviors

Repulsion Attraction Parallel orientation Searching Ranges of the basic behavior patterns How to integrate and evaluate information

from different neighbors?

Decision models Averaging models

5

15/11 - 05 13 Emergent Systems, Jonny Pettersson, UmU

Huth and Wissel (1992) Model – Limitations

No addressing of external influences No obstacle avoidance No avoidance behaviors such as:

Flash expansion Fountain effect

Recent work (1997 – 2000) has addressed

some of these issues

15/11 - 05 14 Emergent Systems, Jonny Pettersson, UmU

Boids

Craig Reynolds Flocks, Herds, and Schools: A Distributed

Behavioral Model

15/11 - 05 15 Emergent Systems, Jonny Pettersson, UmU

Boids – Separation

6

15/11 - 05 16 Emergent Systems, Jonny Pettersson, UmU

Boids - Alignment

15/11 - 05 17 Emergent Systems, Jonny Pettersson, UmU

Boids - Cohesion

15/11 - 05 18 Emergent Systems, Jonny Pettersson, UmU

Boids - Neighborhood

7

15/11 - 05 19 Emergent Systems, Jonny Pettersson, UmU

Summary

Schooling of fish Boids

15/11 - 05 20 Emergent Systems, Jonny Pettersson, UmU

Next time

Swarm algorithms