FlightEnvVec
The wrapper of Flightmare.
- class FlightEnvVec(impl)
An OpenAIGym like environment of Flightmare.
- env = FlightEnvVec.FlightEnvVec(QuadrotorEnv_v1(
dump(cfg, Dumper=RoundTripDumper), False))
- param impl:
Implementation of flightgym
- type impl:
flightgym.QuadrotorEnv_v1
- seed(self, seed=0)
Set a random seed for the environment.
- Parameters:
seed (int) – Random seed
- Return type:
bool
- step(self, action)
Simulate one step with the given thrusts.
Get observations and the total reward of this step.
// thrusts of each propelleraction = [f1, f2, f3, f4]//full states of the quadrotorobs = [position, orientation, linear velocity, angular velocity]- Parameters:
action (np.ndarray) – thrusts of each propeller
- Return type:
observations, reward, done, info
Error
stepUnity() not yet implemented.
- stepUnity(self, action, send_id)
Simulate one step with the given thrusts and render it in Unity.
Get observations and the total reward of this step.
// thrusts of each propelleraction = [f1, f2, f3, f4]//full states of the quadrotorobs = [position, orientation, linear velocity, angular velocity]- Parameters:
action (np.ndarray) – thrusts of each propeller
send_id (int) – frame ID
- Return type:
observations, reward, done, info
- sample_actions(self)
Sample an action.
// thrusts of each propelleraction = [f1, f2, f3, f4]- Return type:
action (np.ndarray)
- reset()
Reset the quadrotor state and get observations.
//full states of the quadrotorobs = [position, orientation, linear velocity, angular velocity]- Return type:
obs
- reset_and_update_info(self)
Reset the quadrotor state, get observations and update info.
//full states of the quadrotorobs = [position, orientation, linear velocity, angular velocity]- Return type:
obs, info
Error
render() not yet implemented.