Last time � Cellular automata � One-dimensional � Wolfram’s classification � Langton’s lambda parameter � Two-dimensional • Conway’s Game of Life � Pattern formation in slime molds � Dictyostelium discoideum � Modeling of pattern 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 1 Outline for today � Self-Organization � Autonomous Agents � Real Ants � Virtual Ants � Ant Algorithms � Assignment 2 � Assignment 3 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 2 Self-Organization � ”Self-organization is a process in which pattern at the global level of a system emerges solely from numerous interactions among the lower-level components of the system. Moreover, the rules specifying interactions among the system’s components are executed using only local information, without reference to the global pattern.” – Camazine et al, p. 8 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 3 1
Self-Organization � Pattern � A particular, organized arrangement of objects in space or time � Interactions � Based on local information only - no global information � Physical laws � Genetically controlled properties of the components 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 4 Self-Organization - Ingredients � Positive feedback � Activity amplification � Negative feedback � Activity balancing � Amplification of random fluctuations � Multiple interactions 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 5 Self-Organization - Information � Signals � Stimuli shaped by natural selection specifically to convey information � Cues � Stimuli that convey information only incidentally � Gathered from one’s neighbors � Stimuli-response, simple behavioral rules of thumb � Gathered from work in progress � Stigmergy � Random fluctuation and chance heterogeneities 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 6 2
Self-Organization - Characteristics � Dynamic systems � Exhibit emergent properties � Attractors � Multistability � Bifurcations � Parameter tuning � Environmental factors � Adaptive systems � Different patterns may result from the same mechanism � Simple rules, complex patterns 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 7 Self-Organization – Alternatives � Central leader � Need effective communication and cognitive abilities � Blueprints � Most be stored � Recipes � Hinders flexibility � Templates � Must be available 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 8 Stigmergy � A recursive control system � Effective for coordination in space and time � A sequence of qualitatively different stimulus-response behaviors � Two types: � Qualitative stigmergy � Quantitative stigmergy 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 9 3
Stigmergy - Advantages � Permits simpler agents � Decrease direct communication between agents � Incremental improvement � Flexible, since when environment changes, agents respond appropriately 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 10 Autonomous Agent � ”a unit that interacts with its environment (which probably consists of other agents) � but acts independently from all other agents in that it does not take commands from some seen or unseen leader, � nor does an agent have some idea of a global plan that it should be following.” - Flake, p. 261 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 11 Real Ants � Imagine if artificial systems could do the things ants can do? � Why ants? � Amazonas: 30% of biomass is ants/termites � Amazonas: dry weight of social insects is four times that of other land animals � Earth: ~10% of total biomass (like humans) 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 12 4
Army Ants � 100 000s in colony � Create temporary ”bivouacs” � Act like unified entity (Pictures from AntColony.org) 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 13 Fungus-Growing Ants � "A Leaf Cutter Colony can strip the tallest of trees in a single day. Equivalent consumption of a full grown cow in the same time!" � ”Cultivate” fungi underground � Fertilize with compost from chewed leaves (Pictures from AntColony.org) 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 14 Fungus Cultivator Nest (Picture from AntColony.org) 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 15 5
Harvester Ants � Find shortest path to food � Prioritize food sources based on distance and ease of access (Picture from The Texas A&M University System) 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 16 Langton’s Virtual Ants � Grid with white or black squares � Virtual ants can face N, S, E, W � Behavioral rule: � Take a step forward � if on a white square then paint it black and turn 90º right � if on a black square then paint it white and turn 90º left 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 17 Virtual Ants - Example 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 18 6
Virtual Ants – Time Reversibility � Virtual ants are time-reversible � But, time-reversibility does not imply global simplicity � Even a single virtual ant interacts with its own prior history � Demonstration 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 19 Virtual Ants - Conclusion � Even simple, reversible local behavior can lead to complex global behavior � Such complex behavior may create structures as well as apparently random behavior 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 20 Ant Algorithms � Ant colony optimization (ACO) � Developed in 1991 by Dorigo (PhD dissertation) in collaboration with Colorni and Maniezzo 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 21 7
Summary � Self-Organization � Autonomous Agents � Real Ants � Virtual Ants � Ant Algorithms � Assignment 2 � Assignment 3 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 22 Next time � Flocks, Herds, and Schools � Boids 11/11 - 05 Emergent Systems, Jonny Pettersson, UmU 23 8
Recommend
More recommend
