Integrating acceleration to get the displacement

Hey

I am trying to integrate acceleration to get the displacement as you can see in the figure below. However, the velocity and displacement are showing exponential decrease. I have used you cumtrapz twice to get the displacement. Anybody know what's the issue here? I also don't want to use detrend to remove the polynomial part,

[Torsten]

The curves for v and z look reasonable. Acceleration is almost always negative, thus velocity decreases throughout and is negative, thus z continuously decreases starting from 0. What's your problem ?

[Sulaymon Eshkabilov]

Can you share your numerical data to verify what you have showed here is accurate or not?

[Sam Chak]

Just want to add that in motion system, when performing the integrations, the initial velocity and the initial displacement are required to obtain an accurate solution. From your images, it appears that you assumed and .

Answer 1

[Ganapathi Subramanian]

Hi Basem,

It is my understanding that you are trying to integrate acceleration to obtain velocity and displacement using ‘cumtrapz’ function. The velocity and displacement graph looks fine. It looks as exponentially decreasing because the acceleration graph is predominantly negative.

You can also add a high pass filter to the acceleration data for baseline correction. But be careful in selecting cut off frequency as higher value leads to loss in valuable information from data.

You can also refer the below file exchange to convert acceleration to velocity and displacement.

https://www.mathworks.com/matlabcentral/fileexchange/69516-convert-acceleration-data-to-velocity-and-displacement

Answer 2

[Oskar Kilgus]

Hi Basem,

if you are sure that the displacement is not meaningfull (because you know the properties of the physical system observed), i would suggest to have a look into the frequency domain. Maybe you could work with a highpass filter to "detrend" your data from low-frequency disturbances as these may cause drifts in integrated signals.