De-noising an accelerometer with autoregression

  • Thread starter Thread starter ramesses
  • Start date Start date
  • Tags Tags
    accelerometer
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
2 replies · 1K views
ramesses
Messages
16
Reaction score
0
Hi
I want to ask about denoising accelerometre with autoregression after filtring high frequency noise. I analysed the noise with Burg method.
To denoise the signal, I made a simulation with autoregression and subtracted it from the noised signal.
Now, I have a problem. The std of the signale become bigger.

What does it mean ? or this is a wrong way to denoise a signal ?
 
Physics news on Phys.org
It's a tough problem. Can you show an example graph or something like that?
 
I'm not sure that I applied correctly the filter.

do I have to simulate the process the auto-regression, then subtract the simulation for the original signal.

or use the transfer function like that : Y(n) =S(n)*1/(1+Σiαn-iY(n-i)) , where S(n) is the original signal ?

Sorry, I'm not very familiar with signal processing