Cellular Automata/Excitable media

From Wikibooks, open books for an open world
Jump to: navigation, search

An introduction to excitable media[edit]

Excitable media are nonlinear dynamic systems known for exhibiting complex behavior that can be observed as pattern formation. They are usually defined by a reaction-diffusion differential equation.

 u_t(\mathbf{r},t) = D \nabla^2 u(\mathbf{r},t) + f(u(\mathbf{r},t))

The diffusion part provides stability and propagation of information, the reactive part provides interesting local dynamics.

A common example of excitable media are prey-predator systems. Such systems are described by a system of differential equations, one function for each of the observed protagonists.

 u_t(\mathbf{r},t) = D_u \nabla^2 u(\mathbf{r},t) + f(u,v)
 v_t(\mathbf{r},t) = D_v \nabla^2 v(\mathbf{r},t) + g(u,v)

We will discuss two different approaches to modeling excitable media. Discretization of differential equations and modeling with cellular automata.

Boundary conditions[edit]

There are different ways to define boundary conditions for the reaction-diffusion equation.

Dirichlet boundary conditions

The value of the function at the boundary is given explicitly u(\mathbf{r_0},t).

Cyclic boundaries

If the initial condition u(x,0) is supposed to be periodic in space, cyclic boundary conditions can be used.

Zero-flux boundary conditions

If zero-flux is expected at the boundary than the component of the functions first derivative normal to the boundary is zero  \vec{n}\cdot \nabla{u}=0 at the boundary. This can be achieved by reflecting function values from the inside over the boundary to the outside.

Discretization of differential equations using the explicit FTCS method[edit]

The traditional method to simulate excitable media is discretization and numerical computation of the governing PDE. First the FTCS (forward-time centered-space method) discretization method is presented. Explicit methods are the simplest and the equations are similar to a cellular automaton, but are inadequate because of stability and convergence problems.

Single PDE[edit]

We will first observe a single PDE describing a single function.

 u_t(\mathbf{r},t) = D \nabla^2 u(\mathbf{r},t) + f(u(\mathbf{r},t))

One-dimensional problem[edit]

In the one dimensional case the space vector becomes a single variable \mathbf{r}=x. The nabla operator becomes \nabla^2=\frac{\partial^2}{\partial{x^2}}.

 \frac{ \partial u(x,t) }{ \partial t } = D \frac{ \partial^2 u(x,t) }{ \partial x^2 }  + f(u(x,t))

The partial differential equation is discretized.

 \frac{ u(x,t+\Delta{t}) - u(x,t) }{ \Delta{t} } =
\quad D \frac{ u(x+\Delta{x},t)-2u(x,t)+u(x-\Delta{x},t) }{ \Delta{x}^2 }
\quad + f(u(x,t))
Forward-time centered-space method

Each finite element at time t+\Delta{t} is calculated from three neighboring elements at time t (see figure at the right).

 u(x,t+\Delta{t}) = u(x,t) + \,
\quad + d( u(x+\Delta x,t)-2u(x,t)+u(x-\Delta x,t) )
\quad + \Delta{t} f(u(x,t))

where the diffusion number is

 d =  D \frac{ \Delta{t} }{ \Delta{x}^2 }

The FTCS method is stable if

 d \leq 1/2
Boundary conditions

For periodic boundaries present values at the left boundary x=0 can be used to compute the future values at the right boundary x=L_x and the other way round.

 u(0,t+\Delta{t}) = u(0,t) + d( u(0+\Delta x,t)-2u(0,t)+u(L_x,t) ) + \Delta{t} f(u(0,t)) \,
 u(L_x,t+\Delta{t}) = u(L_x,t) + d( u(0,t)-2u(L_x,t)+u(L_x-\Delta x,t) ) + \Delta{t} f(u(L_x,t)) \,

If there is zero-flux \nabla u(0,t)=0 at the boundaries than values outside the boundary are reflections of values inside u(0+\Delta x,t)=u(0-\Delta x,t).

 u(0,t+\Delta{t}) = u(0,t) + 2d( u(0+\Delta x,t)-u(0,t) ) + \Delta{t} f(u(0,t)) \,
 u(L_x,t+\Delta{t}) = u(L_x,t) + 2d( u(L_x,t)+u(L_x-\Delta x,t) ) + \Delta{t} f(u(L_x,t)) \,

Two-dimensional problem[edit]

In the two-dimensional case, the space vector becomes a variable pair \mathbf{r}=[x,y]. The nabla operator becomes \nabla^2=\frac{\partial^2}{\partial{x^2}}+\frac{\partial^2}{\partial{y^2}}.

 \frac{ \partial u(x,y,t) }{ \partial t } = D \left( \frac{ \partial^2 u(x,y,t) }{ \partial x^2 } + \frac{ \partial^2 u(x,y,t) }{ \partial y^2 } \right) + f(u(x,y,t))

The partial differential equation is discretized using the forward-time centered-space method.

\frac{ u(x,y,t+\Delta{t}) - u(x,y,t) }{ \Delta{t} } =
\quad = D \frac{ u(x+\Delta{x},y,t)-2u(x,y,t)+u(x-\Delta{x},y,t) }{ \Delta{x}^2 } +
\quad + D \frac{ u(x,y+\Delta{y},t)-2u(x,y,t)+u(x,y-\Delta{y},t) }{ \Delta{y}^2 } +
\quad + f(u(x,y,t))
Forward-time centered-space method

Each finite element at time t+\Delta{t} is calculated from five neighboring elements at time t (see figure at the right).

 u(x,y,t+\Delta{t}) = u(x,t) + \,
\quad + d_x( u(x+\Delta{x},y,t)-2u(x,y,t)+u(x-\Delta{x},y,t) ) +
\quad + d_y( u(x,y+\Delta{y},t)-2u(x,y,t)+u(x,y-\Delta{y},t) ) +
\quad + \Delta{t} f(u(x,y,t))

where the diffusion numbers are

 d_x=D\frac{\Delta{t}}{\Delta{x}^2} \quad d_y=D\frac{\Delta{t}}{\Delta{y}^2}

The FTCS method is stable if

 d_x \leq 1/2 and  d_y \leq 1/2
Boundary conditions

The same ideas as in the one dimensional case can be used for two dimensions.

System of PDE[edit]

A system of PDE describes two functions that interact with each other (prey-predator).

 u_t(\mathbf{r},t) = D_u \nabla^2 u(\mathbf{r},t) + f(u,v)
 v_t(\mathbf{r},t) = D_v \nabla^2 v(\mathbf{r},t) + g(u,v)

The interaction is local, which means, the dispersion part can be computed separately for each equation, and than the reaction part is added to the result.

Other PDE discretization methods[edit]

Modeling with cellular automata[edit]

Greenberg-Hastings Model[edit]



  1. Joe D. Hoffman, Numerical Methods for Engineers and Scientists
  2. Toffoli T., Margolus N., Cellular Automata Machines: A New Environment for Modeling, The MIT Press (1987), Cambridge, Massachusetts
  3. Toffoli T., Cellular automata as an alternative to Differential equations, in Modeling Physics, Physica 10D, (1984)
  4. http://www.jweimar.de/paper-abstracts.html
  5. Robert Fisch, Janko Gravner, David Griffeath, Threshold-Range Scaling of Excitable Cellular Automata
  6. Robert Fisch, Janko Gravner, David Griffeath, Metastability in the Greenberg-Hastings Model
  7. Marcus R. Garvie Finite difference schemes for reaction-diffusion equations modeling predator-prey interactions in MATLAB