LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

kalman implementation in C

If you store the state internal to the VI (shift register or Feedback node) you do have to make the VI reentrant. Otherwise the storage will indeed be interdependant since it is in fact the same. That has nothing to do with the Call DLL Node.

 

Personally I would rather maintain the storage for the state in the caller in a shift register and pass it to the filter VI in each iteration and store the modified state back into the shift register after each execution. Storing the state internal to the VI has its advantages but also makes it all a bit non-intuitive when looking at the code. Also the diagram with the Feedback node and all the extra stuff needed to allow initializing it simply doesn't look nice to me.

Rolf Kalbermatter
My Blog
Message 11 of 11
(585 Views)