Merge pull request #158 from UoA-CARES/chore/clean-up

Chore/clean up

src/environments/environments/CarTrackEnvironment.py

@@ -1,11 +1,12 @@
 import math
 import rclpy
+import numpy as np
 from rclpy import Future
 import random
 from environment_interfaces.srv import Reset
 from environments.F1tenthEnvironment import F1tenthEnvironment
 from .termination import has_collided, has_flipped_over
-from .util import process_odom, avg_lidar, create_lidar_msg, get_all_goals_and_waypoints_in_multi_tracks
+from .util import process_odom, avg_lidar, create_lidar_msg, get_all_goals_and_waypoints_in_multi_tracks, ackermann_to_twist
 from .goal_positions import goal_positions
 from .waypoints import waypoints
 from std_srvs.srv import SetBool
@@ -133,7 +134,11 @@ def step(self, action):
         self.call_step(pause=False)
         _, full_state = self.get_observation()
 
-        lin_vel, ang_vel = action
+        lin_vel, steering_angle = action
+
+        # TODO: get rid of hard coded wheelbase
+        ang_vel = ackermann_to_twist(steering_angle, lin_vel, 0.315)
+
         self.set_velocity(lin_vel, ang_vel)
 
         self.sleep()
@@ -162,8 +167,6 @@ def get_observation(self):
         num_points = self.LIDAR_POINTS
 
         reduced_range = avg_lidar(lidar, num_points)
-
-        # Get Goal Position
 
         match (self.observation_mode):
             case 'no_position':

src/reinforcement_learning/config/sanity_check.yaml

@@ -1,18 +1,4 @@
 sanity_check:
   ros__parameters:
-    environment: 'CarTrack' # CarGoal, CarWall, CarBlock, CarTrack
-    track: 'austin_track' # track_1, track_2, track_3 -> only applies for CarTrack
-    max_steps_exploration: 5000
-    max_steps_training: 1000000
-    reward_range: 1.0
-    collision_range: 0.2
-    # gamma: 0.95
-    # tau: 0.005
-    # g: 5
-    # batch_size: 32
-    # buffer_size: 1000000
-    # actor_lr: 0.0001
-    # critic_lr: 0.001
-    # max_steps: 10
-    # step_length: 0.25
-    # seed: 123
+    environment: 'CarGoal' # CarGoal, CarWall, CarBlock, CarTrack
+

src/reinforcement_learning/config/test.yaml

@@ -2,12 +2,12 @@ test:
   ros__parameters:
     environment: 'CarTrack'
     track: 'multi_track' # track_1, track_2, track_3, multi_track, multi_track_testing -> only applies for CarTrack
-    number_eval_episodes: 100
+    number_eval_episodes: 3
     max_steps: 3000
-    actor_path: training_logs/SAC-CarTrack-23_10_25_10:41:10/models/SAC-checkpoint-15000_actor.pht
-    critic_path: training_logs/SAC-CarTrack-23_10_25_10:41:10/models/SAC-checkpoint-15000_critic.pht
-    algorithm: 'SAC'
+    actor_path: training_logs/TD3-CarTrack-24_03_27_16:02:09/models/TD3_actor.pht
+    critic_path: training_logs/TD3-CarTrack-24_03_27_16:02:09/models/TD3_critic.pht
+    algorithm: 'TD3'
     step_length: 0.1
-    reward_range: 3.0
+    reward_range: 3
     collision_range: 0.2
     observation_mode: 'lidar_only'

src/reinforcement_learning/config/train.yaml

@@ -5,5 +5,5 @@ train:
     car_name: 'f1tenth'
     algorithm: 'TD3'
     g: 1
-    buffer_size: 50000
+    buffer_size: 1000000
     observation_mode: 'lidar_only'

src/reinforcement_learning/launch/sanity_check.launch.py

@@ -31,16 +31,7 @@ def generate_launch_description():
         launch_description_source=PythonLaunchDescriptionSource(
             os.path.join(pkg_environments, f'{env_launch[env]}.launch.py')),
         launch_arguments={
-            'track': TextSubstitution(text=str(config['sanity_check']['ros__parameters']['track'])),
-        }.items() #TODO: this doesn't do anything
-    )
-
-    f1tenth = IncludeLaunchDescription(
-        launch_description_source=PythonLaunchDescriptionSource(
-            os.path.join(pkg_f1tenth_bringup, 'simulation_bringup.launch.py')),
-        launch_arguments={
-            'name': 'f1tenth',
-            'world': 'empty'
+            'car_name': TextSubstitution(text=str(config['sanity_check']['ros__parameters']['car_name'] if 'car_name' in config['sanity_check']['ros__parameters'] else 'f1tenth')),
         }.items()
     )
 
@@ -61,6 +52,5 @@ def generate_launch_description():
         SetEnvironmentVariable(name='GZ_SIM_RESOURCE_PATH', value=pkg_f1tenth_description[:-19]),
         SetParameter(name='use_sim_time', value=True),
         environment,
-        f1tenth,
         main
 ])