Workshop on Chaos Fractals and Power Laws Clint

  • Slides: 33
Download presentation
Workshop on. Chaos, Fractals, and Power Laws Clint Sprott (workshop leader) Department of Physics

Workshop on. Chaos, Fractals, and Power Laws Clint Sprott (workshop leader) Department of Physics University of Wisconsin - Madison Presented at the Annual Meeting of the Society for Chaos Theory in Psychology and Life Sciences at Marquette University in Milwaukee, WI on July 31, 2014

Introductions n n n Name? Affiliation? Field? Level of expertise? Main interest? u Chaos

Introductions n n n Name? Affiliation? Field? Level of expertise? Main interest? u Chaos u Fractals u Power laws

Connections Chaos Fractals Chaos makes fractals Fractals are the “fingerprints of chaos” Fractals obey

Connections Chaos Fractals Chaos makes fractals Fractals are the “fingerprints of chaos” Fractals obey power laws Power Laws The power is the dimension of the fractal

Dynamical Systems Deterministic Linear Transient Stochastic (Random) Nonlinear Periodic Quasiperiodic Chaotic

Dynamical Systems Deterministic Linear Transient Stochastic (Random) Nonlinear Periodic Quasiperiodic Chaotic

Chaos n Sensitive dependence on initial conditions n Topologically mixing n Dense periodic orbits

Chaos n Sensitive dependence on initial conditions n Topologically mixing n Dense periodic orbits

Heirarchy of Dynamical Behaviors n n n Regular predictable (clocks, planets, tides) Regular unpredictable

Heirarchy of Dynamical Behaviors n n n Regular predictable (clocks, planets, tides) Regular unpredictable (coin toss) Transient chaos (pinball machine) Intermittent chaos (logistic map, A = 3. 83) Narrow band chaos (Rössler system) Broad-band low-D chaos (Lorenz system) Broad-band high-D chaos (ANNs) Correlated (colored) noise (random walk) Pseudo-randomness (computer RNG) Random noise (radioactivity, radio ‘static’) Combination of the above (most real-world phenomena)

Chaotic Systems n Discrete-time (iterated maps) / continuous time (ODEs) n Conservative / dissipative

Chaotic Systems n Discrete-time (iterated maps) / continuous time (ODEs) n Conservative / dissipative n Autonomous / non-autonomous n Chaotic / hyperchaotic n Regular / spatiotemporal chaos (cellular automata, PDEs)

Bifurcation Diagram for Chaotic Circuit

Bifurcation Diagram for Chaotic Circuit

Stretching and Folding

Stretching and Folding

Lyapunov Exponents 1 = <log(ΔRn/ΔR 0)> / Δt

Lyapunov Exponents 1 = <log(ΔRn/ΔR 0)> / Δt

Other Chaos Topics n n n n n Limit cycles Quasiperiodicity and tori Poincaré

Other Chaos Topics n n n n n Limit cycles Quasiperiodicity and tori Poincaré sections Transient chaos Intermittency Basins of attraction Bifurcations Routes to chaos Hidden attractors

Fractals n n n Geometrical objects generally with non-integer dimension Self-similarity (contains infinite copies

Fractals n n n Geometrical objects generally with non-integer dimension Self-similarity (contains infinite copies of itself) Structure on all scales (detail persists when zoomed arbitrarily)

Fractal Types n n Deterministic / random Exact self-similarity / statistical self-similarity n Self-similar

Fractal Types n n Deterministic / random Exact self-similarity / statistical self-similarity n Self-similar / self-affine n Fractal / prefractal n Mathematical / natural

Cantor Set D = log 2 / log 3 = 0. 6309…

Cantor Set D = log 2 / log 3 = 0. 6309…

Cantor Curtains

Cantor Curtains

Fractal Curves

Fractal Curves

Weisstrass Function

Weisstrass Function

Fractal Trees

Fractal Trees

Lindenmayer Systems

Lindenmayer Systems

Fractal Gaskets

Fractal Gaskets

Natural Fractals

Natural Fractals

Fractal Dimension

Fractal Dimension

Other Fractal Topics n n n n Julia sets Diffusion-limited aggregation Fractal landscapes Multifractals

Other Fractal Topics n n n n Julia sets Diffusion-limited aggregation Fractal landscapes Multifractals Rényi (generalized) dimensions Iterated function systems Cellular automata Lindenmayer systems

Power Laws n n n y = xα log y = α log x

Power Laws n n n y = xα log y = α log x α is the slope of the curve log y versus log x Note that the integral of y from zero to infinity is infinite (not normalizable) Thus no probability distribution can be a true power law

Other Properties n No mean or standard deviation n Scale invariant n “Fat tail”

Other Properties n No mean or standard deviation n Scale invariant n “Fat tail”

Power Laws (Zipf) Words in English Text Size of Power Outages Earthquake Magnitudes Internet

Power Laws (Zipf) Words in English Text Size of Power Outages Earthquake Magnitudes Internet Document Accesses

Other Examples of Power Laws n n n Populations of cities Size of moon

Other Examples of Power Laws n n n Populations of cities Size of moon craters Size of solar flares Size of computer files Casualties in wars Occurrence of personal names Number of papers scientists write Number of citations received Sales of books, music, … Individual wealth, personal income Many others …

References n n n http: //sprott. physics. wisc. edu/ lectures/sctpls 14. pptx (this talk)

References n n n http: //sprott. physics. wisc. edu/ lectures/sctpls 14. pptx (this talk) http: //sprott. physics. wisc. edu/chaost sa/ (my chaos textbook) sprott@physics. wisc. edu (contact me)