Hi there,
I’ve been working on the motion planning problem with only statistic obstacles. During the Tutorial 2 Vanilla Learning, I received the Attribute Error:Velocity, pointed to
gym_commonroad/observation/surrounding_observation.py in get_rel_v_p_follow_leading(distance_sign, distance_abs, p_rel_follow, p_rel_lead, v_rel_follow, v_rel_lead, obs_state, obstacle, ego_state, o_follow, o_lead, obstacle_follow, obstacle_lead)
767 ego_state.velocity_y, ego_state.velocity)
768 delta_orientation = obs_state.orientation - ego_state_orientation
--> 769 v_rel_lead = obs_state.velocity * np.cos(delta_orientation) - ego_state.velocity
770 p_rel_lead = distance_abs
771 o_lead = obs_state
I tried to set the obs and ego velocity before it, but the result is still the same. The difference between tutorial scenarios with mine, is that tutorial scenarios have multiple dynamic obs,but mine only static obstacles. Also for my case, I modify the obs_id into 3 digits. Does the RL method only work for dynamic obstacles now? Or is there other rules I haven’t obey when generating the xmls?
BR
Joanne