LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Help! Robot Manipulator model does not work. Can't see why!

Hi guys. 

I am working on a school project where I am modeling and making a control system for a robot manipulator. After doing all the math, and implementing a working example in simulink, I tried to 'copy' the dynamic model intro Labview. Unfortunately, this was a lot harder than I though. I am using Simulation Subsystem and continuous functions, like integrators, just like I do in Simulink. Since the matrices in this model is quite big (enormous), I decided to use mathscript where I implement vectors and matrices that are depending on joint position and velocity. 

All in all, the problem is that the simulation in labview does not give the same result as the simulation in Simulink. Is is supposed to settle at some specified degrees when no external force is applied.  

Please help me guys. I have been working on this for way to long.

0 Kudos
Message 1 of 4
(2,164 Views)

Hello

 

In order to help you, please attach all of the subVIs that you created for your model or create a source distribution (http://zone.ni.com/reference/en-XX/help/371361H-01/lvhowto/build_source_distrib/).

 

Also, I noticed that your top level VI (in attachment on your post) is using a timed-loop with a Simulation Subsystem. In general, we would use a simulation loop instead. Why did you use a timed-loop? Also, which solver did you use on the other software? The way you setup, you have to use fixed step size which is several cases would give different results.

Barp - Control, Simulation, RTT and HIL - National Instruments
Message 2 of 4
(2,146 Views)

Hi.

I have now attached an LLB-file. Hope that helps:)


I used a timed loop in order to run the simulation several times. I also use the time when making a plot of the simulated joint positions. I did not now that it is preferred to use a simulation loop with a simulation subsystem. I though that was too much.  I would like to use a solver for stiff systems, but I was not able to choose any other than Runga-Kutta 1 to 4. 

I tried to skip the timed loop, and adding a simulation loop to the VI. This did not give the desired/expected response, but at least I didn't get INF as I used to! Hope you can help me even more:)

Morten

0 Kudos
Message 3 of 4
(2,138 Views)

Hi again. 

I'm sorry for nagging, but I really hope that someone can help me. I have no idea whats wrong and I am quite new to this software. 

Please give me a clue if you can!

Morten

0 Kudos
Message 4 of 4
(2,125 Views)