The measurement of force sensors in Stewart platform

[biscuit122]

I constructed a Stewart platform in Vrep demonstrated in Fig.1 and output the measurement values from force sensors. The measured force is supported by 6 electric cylinders to top platform. Why does the curves appear so chaotic? The Fig.2 is the output from Vrep and Fig.3 is from adams. Their algorithms are the same with each other.

[coppelia]

Hello,

did you try to switch to a different engine? Did you try to adjust the dynamics step size?
Also, remember that during one simulation step, you normally have (by default) 10 dynamics steps. So reading the force sensor is basically the results of 10 averaged measurements. The simplest test would be to set the simulation step size to the same as the dynamics step size.

Cheers

[biscuit122]

Hello,

did you try to switch to a different engine? Did you try to adjust the dynamics step size?
Also, remember that during one simulation step, you normally have (by default) 10 dynamics steps. So reading the force sensor is basically the results of 10 averaged measurements. The simplest test would be to set the simulation step size to the same as the dynamics step size.

Cheers

Thank you first. Actually, I have attempted to switch to different simulation step size and engine. But the result become more terrible. The result I showed is the best one. So is it the normal in V-rep?

[coppelia]

It is not directly computed by V-REP, but by the used physics engine (i.e. Bullet, ODE, Vortex or Newton). It also of course depends on your mass/inertia settings. Can you post your scene?

Cheers

[biscuit122]

It is not directly computed by V-REP, but by the used physics engine (i.e. Bullet, ODE, Vortex or Newton). It also of course depends on your mass/inertia settings. Can you post your scene?

Cheers

Thank you so much. I have post my scence in this problem viewtopic.php?f=9&t=7308
And I have not direct any disired motion command to it but the program is still operating. I speculate the reason for these two probems may be the same.