Analyzing vibration using machine learning.

Thread Starter

goutham1995

Joined Feb 18, 2018
104
  1. Hi,
I would like a few suggestions on an idea that I have -

I am trying to make a musical instrument (percussion), whilst just having a PVC disc. I am hitting the disc in a variety of styles (in order to produce a variety of sounds correspondingly) , just like the way the actual percussion instrument is hit. I am converting the mechanical vibrations on the PVC disc to an electrical signal using a transducer, performing an FFT analysis of the different strokes, and trying to identify the stroke which is hit. Using this technique, I could get an accuracy of only 80 percent. I would like it to be extremely accurate ( more than 95 percent recognition). I was using only frequency as the parameter used to distinguish the sounds.

Now, I am thinking that if I could use other parameters too in order to identify the stroke, I might be able to get the required accuracy. I am thinking of resorting to Machine Learning for this. I am kind of new to this and would like to know what I might need to know before I proceed with this idea.

Any help would be greatly appreciated.
 

KeithWalker

Joined Jul 10, 2017
3,050
  1. Hi,
I would like a few suggestions on an idea that I have -

I am trying to make a musical instrument (percussion), whilst just having a PVC disc. I am hitting the disc in a variety of styles (in order to produce a variety of sounds correspondingly) , just like the way the actual percussion instrument is hit. I am converting the mechanical vibrations on the PVC disc to an electrical signal using a transducer, performing an FFT analysis of the different strokes, and trying to identify the stroke which is hit. Using this technique, I could get an accuracy of only 80 percent. I would like it to be extremely accurate ( more than 95 percent recognition). I was using only frequency as the parameter used to distinguish the sounds.

Now, I am thinking that if I could use other parameters too in order to identify the stroke, I might be able to get the required accuracy. I am thinking of resorting to Machine Learning for this. I am kind of new to this and would like to know what I might need to know before I proceed with this idea.

Any help would be greatly appreciated.
You are on the right track using FFTs to display the differences, I suggest that you try using envelope fitting. If you don't recognize the term, read up on it. There are lots of good articles on-line. Once you are familiar with the principal, you can use machine learning to compare envelopes while you adjust the parameters and limits for each type of signal.
The way to do it is by saving the envelope of a particular signal, and then extending different parts of the envelope with subsequent envelopes of the same signal until you have a reliable "pass/fail" limit. Repeat this for each of the different signals.
Regards,
Keith
 
Top