vikashplus / franka_sim Goto Github PK

Hello,
Thanks very much for the repo, I am also building a mujoco simulation for Franka Panda arm. However, when I use the actuator settings in your model file, the behavior of the arm is problematic, basically, if I obtain the current position of the joint, and then send it again to the robot, the actuators force will jump to smaller values and then starts to move to its value while the robot stands still, here is the snippet of my code:

`id_1 = e.env.sim.model.joint_name2id('panda_finger_joint1')
id_2 = e.env.sim.model.joint_name2id('panda_finger_joint2')
# print(id_1, id_2)
# qp[7] -= 0.02
# qp[8] -= 0.02
qp_curr = qp.copy()

    for i in range(200):
        print(i)
        qp_curr[7] = qp_curr[id_1]
        qp_curr[8] = qp_curr[id_2]
        e.env.move_to_jnt(qp_curr)
        qp, _, _, _ = e.env.get_env_state()
        id_1 = e.env.sim.model.joint_name2id('panda_finger_joint1')
        id_2 = e.env.sim.model.joint_name2id('panda_finger_joint2')
        # print(id_1, id_2)
        qp_curr = qp.copy()`

and in the following image you can see the sudden "jump" of the actuator force, which will cause the arm to move:

Do you have any idea how to fix this ?