I want to get displacement from accelerometric sensors, but through analysis data, it has a big difference with the fact. I want to know if there is anyone, who did research about this and can give me some advise.
Look on the physiotherapychoices.org.au website as that is an evidence based website so maybe there is some research there.
The problem is the noise, in order to reduce the error you must acquire as more samples per second than you can and increase the ADC bits/sample to at least 14 bits.
To reduce the noise you should mount the acceleration sensors on neoprene rubber that should reduce vibrations. You can use a low pass filter to reduce the noise further.
It is not possible to predict the error due the noise is unpredictable BUT you can have an idea of it by acquiring as example at a data rate of 100KSPS and obtaining the position and compare result by using 50ksps and interpolate.
I recommend use a high speed accelerometer the more linear with the frequency as possible.
I hope its helps
- Badges
- Science topic
- Similar topics
- Methodology
- Sensors