I am trying to simulate a robot arm in pyfmi using a FMU created in maplesim. The simulation is in real time with inputs and outputs that are changed each cycle.
I therefor have taken a lott of inspiration from the fmi example bouncing_ball_native that comes with the pyfmi package.
I am using a FMI version 2.0 type ME.
However when I use the FMU model in pyfmi the z axis is inverted, it do not follow the right hand rule. When the robot arms end effector is moving in positive direction along the z-axis then the model says that the end effector is moving in negative direction along the z-axis, mirrored.
This model is based on a real robot so I have been able to compare its reactions to position control quiet well. The model tested inside of maplesim do react almost exactly as the real robot with the TCP (toll center point) of the simulated and real robot being the same. There is no sign of an inverted z-axis in the maplesim model. As stated previously the end effector coordinates of the real and simulated robot arm are the same inside of maplesim.
However during use in pyfmi, something happens and the z-axis is inverted. I have no idea what the problem is as there are no code or parameters used in the using of pyfmi that have any impact on the position of the z-direction.
Have any of you experienced and knowledgeable people ever experienced something like this and could know what is wrong?
With the greatest of thanks Eric
Update! After testing around have I found out that the inverser z-direction is caused by the changing of the directions on joint 2 and 5. So instead of pyfmi changing my z-axis for no reason is it changing the directions of two of my joints.
I am still as confused but maybe it will help you.
