Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Thread Subject:
Accelerometer output - Kalman filter

Subject: Accelerometer output - Kalman filter

From: smoindrot@aol.com (Sebastien Moindrot)

Date: 2 Apr, 2001 10:32:02

Message: 1 of 4

Hi there,

I have little knowledge with Kalman filters, but I think I understand
the abouts of it. I have accelerometers measuring raw data and I
would like to use a Kalman filter to get a better data to integrate
for the velocity and then the position.
I have seen some codes for filters but I don't really understand the
inputs:
[x, V, VV, loglik] = kalman_filter(y, A, C, Q, R, init_x, init_V,
model).(http://www.cs.berkeley.edu/~murphyk/Bayes/kalman.html)
I have a vector acceleration, time sample (or time vector) and this is
it.
Could you please help me.

Cheers,
Sebastien

Subject: Accelerometer output - Kalman filter

From: Omur Bas

Date: 2 Apr, 2001 11:19:47

Message: 2 of 4

If you post the m file for kalman_filter, it'll be much easier to
help you, but here's my guess: You have a model
x(k+1)=Ax(k)+p(k)
y(k)=Cx(k)+m(k)
p(k) is process noise and m(k) is measurement noise. With this
model, the function inputs should be:
y: acceleration measurements
A: A matrix in the model
C: C matrix in the model
init_x: Initial position (x1(0) in the model)
init_v: Initial velocity (x2(0) in the model)
Q: Covariance matrix for p(k)
R: covariance matrix for m(k)

Hope this helps!

Omur

Sebastien Moindrot <smoindrot@aol.com> wrote in message
news:wehj160yitxf@forum.mathforum.com...
> Hi there,
>
> I have little knowledge with Kalman filters, but I think I understand
> the abouts of it. I have accelerometers measuring raw data and I
> would like to use a Kalman filter to get a better data to integrate
> for the velocity and then the position.
> I have seen some codes for filters but I don't really understand the
> inputs:
> [x, V, VV, loglik] = kalman_filter(y, A, C, Q, R, init_x, init_V,
> model).(http://www.cs.berkeley.edu/~murphyk/Bayes/kalman.html)
> I have a vector acceleration, time sample (or time vector) and this is
> it.
> Could you please help me.
>
> Cheers,
> Sebastien

Subject: Accelerometer output - Kalman filter

From: Johan Kullstam

Date: 3 Apr, 2001 05:11:26

Message: 3 of 4

smoindrot@aol.com (Sebastien Moindrot) writes:

> Hi there,
>
> I have little knowledge with Kalman filters, but I think I understand
> the abouts of it. I have accelerometers measuring raw data and I
> would like to use a Kalman filter to get a better data to integrate
> for the velocity and then the position.

usually the point of the kalman filter is to integrate up the
accelerometers and give you velocity and position.

what you do is make a model of the body the accelerometer is atached
to and the likely inputs. assume it's linear. (non-linear models are
tougher, but you linearize up front or use extended kalman filter to
linearize as you go.) express it in discrete time state space form

x(t+1) = A * x(t) + B * u(t) + w(t)
y(t) = C * x(t) + D * u(t) + v(t)

where u is the known driving input
x is the system state
y is the measurement
w is a noise driving the plant (unknown inputs)
v is a measurement noise

you need the model for the motion or the kalman filter won't help you.

> I have seen some codes for filters but I don't really understand the
> inputs:
> [x, V, VV, loglik] = kalman_filter(y, A, C, Q, R, init_x, init_V,
> model).(http://www.cs.berkeley.edu/~murphyk/Bayes/kalman.html)
> I have a vector acceleration, time sample (or time vector) and this is
> it.
> Could you please help me.
>
> Cheers,
> Sebastien

--
J o h a n K u l l s t a m
[kullstam@ne.mediaone.net]
Don't Fear the Penguin!

Subject: Accelerometer output - Kalman filter

From: John Lukesh

Date: 4 Apr, 2001 04:02:02

Message: 4 of 4

I recommend the book by Robert Grover Brown and Patrick Y. C. Hwang,
"Introduction to Randon Signals and Applied Kalman filtering". This has a
chapter on inertial navigation and includes MATLAB exercises.

Regards, John

Sebastien Moindrot wrote:

> Hi there,
>
> I have little knowledge with Kalman filters, but I think I understand
> the abouts of it. I have accelerometers measuring raw data and I
> would like to use a Kalman filter to get a better data to integrate
> for the velocity and then the position.
> I have seen some codes for filters but I don't really understand the
> inputs:
> [x, V, VV, loglik] = kalman_filter(y, A, C, Q, R, init_x, init_V,
> model).(http://www.cs.berkeley.edu/~murphyk/Bayes/kalman.html)
> I have a vector acceleration, time sample (or time vector) and this is
> it.
> Could you please help me.
>
> Cheers,
> Sebastien

Tags for this Thread

No tags are associated with this thread.

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us