Optimize parameter of Simulink model to fit experimetal data

Hello everyone,

I'm simulating a vehicle's dynamics and I would like to optimize some of the parameters of my model in order to fit data collected experimetally. The model's differential equations are defined inside of a Matlab function block (as in the attached figure), the time depending inputs of the model are the inputs of the real vehicle except for the longitudinal forces where a closed-loop on the longitudinal velocity is done. I'd like to find some constant parameters (such as cornering stiffness) of the model, is it possible to use an approach that is similar to fmincon function? How can I find a solution to such problem?

Thank you,

Juan

Answer 1

[Alan Weiss]

I am not sure, but maybe the approaches in one of these examples will help:

lsqnonlin with a Simulink® Model

Using fminimax with a Simulink® Model

Alan Weiss

MATLAB mathematical toolbox documentation