google-deepmind / mujoco_mpc

Real-time behaviour synthesis with MuJoCo, using Predictive Control
https://github.com/deepmind/mujoco_mpc
Apache License 2.0
1.02k stars 151 forks source link

Modifications on Dynamics, Optimiser of Quadrotor #340

Closed lakshmikanth-nageswar closed 2 months ago

lakshmikanth-nageswar commented 2 months ago

Hi,

Could you please tell where the dynamics of the different models [needed for quadrotor specifically] is defined. I am intending to modify the dynamics and looking to implement different controllers in the environment. I am looking forward for implementation in python. Could someone guide me on the same?

Thanks!

yuvaltassa commented 2 months ago

MuJoCo is the dynamics. The physics code lives here https://github.com/google-deepmind/mujoco/tree/main/src/engine

lakshmikanth-nageswar commented 2 months ago

MuJoCo is the dynamics. The physics code lives here https://github.com/google-deepmind/mujoco/tree/main/src/engine

Thanks for directly to the code. I am looking to modify the constraints, dynamics and implement a different controller in Python for the quadrotor. If I am intending to implement my own controller on the linear dynamics scripted externally, then how to read state data and input control commands from the explicit controller to the environment? Could you please guide a little on that?

Thanks!

yuvaltassa commented 2 months ago

MuJoCo is open-source and well-documented https://mujoco.readthedocs.io/en/latest/overview.html

Read the docs, change the code however you like 🙂