simsd - Simulate models with uncertainty using Monte Carlo method
Syntax
simsd(m,u)
simsd(m,u,N,'noise',Ky)
[y,ysd] = simsd(m,u)
Description
u is an iddata object
containing the inputs. m is a model given as any idmodel object. N random
models are created according to the covariance information given in m.
The responses of each of these models to the input u are
computed and graphed in the same diagram. If the argument 'noise' is
included, noise is added to the simulation in accordance with the
noise model of m and its own uncertainty. Ky denotes
the output numbers to be plotted. (The default is all).
The default value is N=10.
With output arguments
[y,ysd] = simsd(m,u)
No plots are produced, but y is returned
as a cell array with the simulated outputs, and ysd is
the estimated standard deviation of y, based on
the N different simulations. If u is
an iddata object, so are the contents of the cells
of y and ysd; otherwise, they
are returned as vectors/matrices. In the iddata case,
plot(y{:})
thus plots all the responses.
sim and simsd have
similar syntaxes. Note that simsd computes the
standard deviation by Monte Carlo simulation, while sim uses
differential approximations (the Gauss approximation formula). They
might give different results.
Examples
Plot the step response of the model m and
evaluate how it varies in view of the model's uncertainty.
step1 = [zeros(5,1); ones(20,1)];
simsd(m,step1)
See Also
 | sim(idnlhw) | | size |  |
Includes the most popular MATLAB recorded presentations with Q&A sessions led by MATLAB experts.
Get the Interactive Kit
Store 
