f

#### Need matlab help - accelerometer problem

```Hi,

I am trying to solve an interesting problem. I have a handheld device that contains a basic 3d accelerometer. I am trying to determine the path that the device moves through the air and the timing associated with the device. Think of the device being moved like a tennis racket or golf club.

Normally this would be a simple problem to solve as I have the acceleration data. THe complication is that the device swivels in the persons hand and so i do not really know which direction gravity is acting.

I am trying to use Matlab to see if i can figure out the path or the timing (even just timing would be useful to me).

I need some help from someone who has a strong  physics and math background to help me.

Anyone interested?
```
 0
2/25/2009 3:20:02 PM
comp.soft-sys.matlab 211265 articles. 25 followers. lunamoonmoon (257) is leader.

1 Replies
751 Views

Similar Articles

[PageSpeed] 38

```"Tony Davis" <tdavis1198@mac.com> wrote in message <go3nj2\$pqs\$1@fred.mathworks.com>...
> Hi,
>
> I am trying to solve an interesting problem. I have a handheld device that contains a basic 3d accelerometer. I am trying to determine the path that the device moves through the air and the timing associated with the device. Think of the device being moved like a tennis racket or golf club.
>
> Normally this would be a simple problem to solve as I have the acceleration data. THe complication is that the device swivels in the persons hand and so i do not really know which direction gravity is acting.
>
> I am trying to use Matlab to see if i can figure out the path or the timing (even just timing would be useful to me).
>
> I need some help from someone who has a strong  physics and math background to help me.
>
> Anyone interested?

Because the earth's gravity is equivalent to such a high rate of acceleration - 32 ft/sec^2 - it is vital that all acceleration data be obtained, either 1) with a platform that is known to be absolutely horizontal and non-rotating, or else 2) obtained in three dimensional form along with very accurate orientation data.  Otherwise you are wasting your time, in my opinion.

Matlab's 'ode' functions are quite capable of solving the necessary differential equations, but their accuracy is heavily dependent on the accelerometer's and the orientation data accuracy.  (My background: degrees in math and physics at UC Berkeley)

Roger Stafford
```
 0
2/25/2009 9:01:03 PM