Problem in modeling a leg-wheel robot

Typically: "How do I... ", "How can I... " questions
Post Reply
zwei371
Posts: 10
Joined: 20 Dec 2017, 02:58

Problem in modeling a leg-wheel robot

Post by zwei371 »

Hello!
Some strange phenomena happened when the simulation of the model was done. To present this problem clearly, a hexapod robot model supplied by V-REP itself was put in the same scene, and the video can be obtained at this link:
https://www.dropbox.com/s/4h0ukjo18em5z ... r.mp4?dl=0
In addition, to simulate the spring in the knee joint, a joint at the knee is set the spring-damper mode. However, when the spring and damper of the joint were set different values, no differences could be observed.
The dynamic model of this robot is shown at this link:
https://www.dropbox.com/s/4h0ukjo18em5z ... r.mp4?dl=0
The dynamic properties of two neighboring dynamic shaps and joints set as spring mode are shown in this figure:
https://www.dropbox.com/s/b5etbsp6aqtc165/11.jpg?dl=0
Look forward for your reply!

coppelia
Site Admin
Posts: 10339
Joined: 14 Dec 2012, 00:25

Re: Problem in modeling a leg-wheel robot

Post by coppelia »

Hello,

did you try to simulate that model with the Vortex engine? That engine is much more tolerant to real-world mass and inertia values. It might help to adjust (e.g. increase) the masses and/or inertias. Also, all masses in the model should not have a mass (or inertia) ratio that is too large (e.g. 1/10 is already quite a lot), when working with the Bullet or ODE engines.

Then, for your spring/damper to operate correctly, you also need to set a joint target velocity, which is a limiting velocity for your spring/damper.

Cheers

zwei371
Posts: 10
Joined: 20 Dec 2017, 02:58

Re: Problem in modeling a leg-wheel robot

Post by zwei371 »

Hello,

Thanks for your reply!

I installed the vortex studio essentials and did some comparison simulation. When the springs were not added on the legs, the simulation with the votex engine was beter than that with the ODE engine in vibration. But when the springs were added, the simulation with vortex engine would not only vibrate but also move, while the one with ODE engine just vibrated. In addition, the ODE engine is better than the vortex one in simulating the spring.

Now, I will try to control the joints and see the effect, thank you!

zwei371
Posts: 10
Joined: 20 Dec 2017, 02:58

Re: Problem in modeling a leg-wheel robot

Post by zwei371 »

Hello,

The vibration and move of the robot when Vortex Engine used can be eliminated by adjusing the target velocity of the joint. Now we find Vortex Engine can work better, so we now decide to choose it in the simulation.

Thanks!

Post Reply