Quick Example

To give a feel of what can be done using the toolbox a quick example of how a discrete equivalent can be found for an affine system on a simplex.

Finding a discrete equivalent

%First define the state space simplex
s.dim=2;
s.vertices={[2;1],[1;2],[1,1]};
s=makesimplex(s);

%Then the system
s.dynamics.A=[0 1; 0 -0.3];
s.dynamics.B=[0;1];
s.dynamics.C=[0;0];

%Now the input polytope (in this case it is a simplex)
s.input.dim=1;
s.input.vertices={1,-1};
s.input=makesimplex(s.input);

%Prepare the simplex for computations
s=preparesimplex(s);

%Find a discrete equivalent
de=finddiscreteeq(s);

%Plot the result
plotdisceq(s,de);

The example above is a model of a block sliding on a 1-D surface with viscous friction. The input is the force applied on the block by e.g. a rod.

Computing a control strategy for a PAHS

%First define the pahs.
%This PAHS is included in the toolbox.

pahs=demopahs2();

%Compute the discrete abstraction
dgabs=finddiscgame(pahs);

%Define the control requirements
req=demoreq2();

%Convert the requirements to a specification on the discrete abstraction
spec=req2spec(pahs,req)

%Find a winning strategy
[win,sol]=finddgsol(dgabs,spec);

%Is there a winning strategy
if(win)
    disp('Winning strategy found!');

    %Compute control laws based on the strategy
    ctcat=sol2ctrl(pahs,dgabs,sol);

    %Parameters for the simulation:

    %Initial location and mode
    l0=req.init{1}.loc;
    m0=req.init{1}.modes(1);

    %Initial state
    x0=mean(horzcat(pahs.mode{m0}.loc{l0}.s.vertices{:}),2);

    %Set up the occurrence of events
    events{1}.time=3;
    events{1}.mode=2;

    %Simulation time
    simtime=1000;

    %Timestep
    step=.5;

    %Simulate the controlled pahs
    [x,t,m,l]=simpahs(pahs,ctcat,x0,l0,m0,simtime,events,step);

    %Plot the results
    plotpahs(pahs,spec,x,t,events,m0);
else
    disp('No winning strategy found!')
end

The above example can be found as example2.m in the demo directory of the toolbox