Interesting issue with integrator blocks

[Rasmoo]

Hey guys. Here is a short description of the issue:
I an integrating an acceleration twice to get the position of a valve spool. The position integral has an upper and lower limit to simulate the physical limits of the spool motion. However, the velocity integral experiences a kind of 'wind-up' when these position limits are hit, causing the velocity to keep building as long as the calculated acceleration is going in the same direction. When the acceleration then changes sign, the velocity will have to unwind before the spool moves again.

Obviously, the velocity must be reset to 0 and held there until the acceleration changes sign. However, this has proved quite difficult to do with the saturation and reset ports. Does anyone have a good idea of how to do this?

Thank you very much for your help.

 

[FeX32]

Try using a Matlab embedded function instead. You can use if statements inside to zero the vel.


"Simplicity is the ultimate sophistication." L. da Vinci
- Gian

 

[GregLocock]

In the real world double integrating an accelerometer signal to get displacement often has serious errors.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Rasmoo]

Thanks for the replies. However, I finally figured out a solution, which is to use the 2nd order integral block, that is apparently made for this situation. Didn't even know that block existed until now.