I was checking the min-max torques values of ASTI. And I realized that the values are dramatically changing by the simulation dt.
When sim dt=0.05, the max value=251 Nm; when sim dt=0.001, the max value= 3818 Nm.
(dynamics dt is always 0.05 ms as recommended.) (Engine:Bullet 2.83) (Coppeliasim Edu, ver 4.9.0 rev 6)
There is an inverse propotional relation. Why would it change though? Does not low simulation dt mean more calculations and therefore more accurate?
Note: I magnified ASTI by isometric scaling about 1.32 so it be as tall as Standing Bill.
Relation between Joint Torques and Simulation dt
Re: Relation between Joint Torques and Simulation dt
Hello,
the built-in position controller is an opaque controller that is meant to be robust in all conditions. It is regulating a target velocity proportional to the error.
Normally, one would write a custom controller that regulates force instead of velocity, and linked to inverse dynamics calculations. In next release there will be such an example
Cheers
the built-in position controller is an opaque controller that is meant to be robust in all conditions. It is regulating a target velocity proportional to the error.
Normally, one would write a custom controller that regulates force instead of velocity, and linked to inverse dynamics calculations. In next release there will be such an example
Cheers