First post + FES controller

Dear members of the list HI!

I am a Ph D candidate and i am very glad that i have joined this list.

I have designed, or at least i think i have :), a Functional
Electrical Stimulation (FES) controller based in Artificial
Intelligence techniques, mostly Fuzzy Logic, for the restoration of
human locomotion in patients that suffer from the Drop-Foot syndrome.
(patients unable to raise their foot when they try to walk)

What i need to do, is to test this controller in a simulation
environment before any clnical use. (Otherwise it could be
dangerous :))

The controller is designed in the Fuzzy Logic Toolbox of Matlab and
my hope was to be able to build a model of the lower limb in Simulink
(Matlab). In this way i could connect the controller with the model
and run the simulations in Simulink.

I would very much appreciate if someone could give me some guidelines
on how to connect the controller with the lower limb model that you
have created in order to run simulations in any possible environment
(Simulink, Anyboby or something else).

Simulink would be ideal for me since i somehow have a solution on
simulating the electrical stimulation of the muscles and i could use
the other toolboxes of Matlab too. But i can use any other
suggestions.

Since i only today discovered your software, i cannot express any
well documented suggestions, but my first impression is that such a
software will for sure solve many many problems. I will have a closer
look and let you know if i have any good suggestions.

By the way Matlab is missing of a toolbox for simulating
musculoskeletal systems so maybe a connection between Anybody
language and Matlab could become something nice?

Best regards

Vassilis.

Hi Vassilis and welcome to the AnyScript group!

> I would very much appreciate if someone could give me some
guidelines
> on how to connect the controller with the lower limb model that
you
> have created in order to run simulations in any possible
environment
> (Simulink, Anyboby or something else).

Well, AnyBody is an inverse dynamics system, so it will compute the
muscle forces when you know the movement and loads. If I understand
your problem correctly, then you need to solve the opposite problem,
namely simulating which movement you will derive from your
controller’s stimulation.

That is essentially a forward dynamics problem. We will probably
include forward dynamics into AnyBody at some time, but it is not
among the first things we are planning.

> Since i only today discovered your software, i cannot express any
> well documented suggestions, but my first impression is that such
a
> software will for sure solve many many problems. I will have a
closer
> look and let you know if i have any good suggestions.
>
> By the way Matlab is missing of a toolbox for simulating
> musculoskeletal systems so maybe a connection between Anybody
> language and Matlab could become something nice?

What you can do with AnyBody is to simulate the movement you would
like to see, and then have the system compute backwards to the
necessary active states of the muscles you are trying to control.
This could be used as a sort of offline identification of the
stimulation pattern you are trying to reach. You could then compare
this with the stimulation proposed by your controller.

I hope this makes sense.

Good luck with the system!
John