Prev: ahhhhh....help!
Next: Creating Complex Structure
From: Joshua on 24 Apr 2010 20:29
I have written a program in Matlab that I now have to move to Simulink. I have done some work with simulink but am a little lost in what the best approach is.

The simulation:

I have a model for a dynamic system, namely an inverted pendulum being solved using ode45 in my matlab code. I am then adding noise to the angle measured and calling this the measurement of the true system.

I then try to match this "true system" with a set of 20 other models of the same system (with the same parameters of the pendulum and cart just different initial conditions). I then try to find which model is a best fit over time and use the state of the model of the best system model to be an estimate of the true state of the system (effectively trying to filter noise). When the states of the model systems gets too far off the measured system new initial conditions near where the system currently is are selected.

So what would be a good method to implement this in Simulink.
specifically being able to reset a system(give it a new set of initial conditions).
 | 
Pages: 1
Prev: ahhhhh....help!
Next: Creating Complex Structure