def scenario_deadend_test(self):
waypoints = [(0, 0, 0), (50, 0, 0), (100, 0, 0)]
self.path = QPMI(waypoints)
self.current = Current(mu=0,
Vmin=0,
Vmax=0,
Vc_init=0,
alpha_init=0,
beta_init=0,
t_step=0)
radius = 25
angles = np.linspace(-90, 90, 10) * np.pi / 180
obstalce_radius = (angles[1] - angles[0]) * radius / 2
for ang1 in angles:
for ang2 in angles:
x = 30 + radius * np.cos(ang1) * np.cos(ang2)
y = radius * np.cos(ang1) * np.sin(ang2)
z = -radius * np.sin(ang1)
self.obstacles.append(Obstacle(obstalce_radius, [x, y, z]))
init_pos = [0, 0, 0]
init_attitude = np.array([
0,
self.path.get_direction_angles(0)[1],
self.path.get_direction_angles(0)[0]
])
initial_state = np.hstack([init_pos, init_attitude])
return initial_state
def scenario_test_path_current(self):
initial_state = self.scenario_test_path()
self.current = Current(mu=0,
Vmin=0.75,
Vmax=0.75,
Vc_init=0.75,
alpha_init=np.pi / 4,
beta_init=np.pi / 6,
t_step=0)
return initial_state
def scenario_test_current(self):
initial_state = self.scenario_test()
self.current = Current(
mu=0,
Vmin=0.75,
Vmax=0.75,
Vc_init=0.75,
alpha_init=np.pi / 4,
beta_init=np.pi / 6,
t_step=0) # Constant velocity current (reproducability for report)
return initial_state
def scenario_test_path(self):
self.n_waypoints = len(test_waypoints)
self.path = QPMI(test_waypoints)
self.current = Current(mu=0,
Vmin=0,
Vmax=0,
Vc_init=0,
alpha_init=0,
beta_init=0,
t_step=0)
init_pos = [0, 0, 0]
init_attitude = np.array([
0,
self.path.get_direction_angles(0)[1],
self.path.get_direction_angles(0)[0]
])
initial_state = np.hstack([init_pos, init_attitude])
return initial_state
def scenario_vertical_test(self):
waypoints = [(0, 0, 0), (50, 0, 0), (100, 0, 0)]
self.path = QPMI(waypoints)
self.current = Current(mu=0,
Vmin=0,
Vmax=0,
Vc_init=0,
alpha_init=0,
beta_init=0,
t_step=0)
self.obstacles = []
for i in range(7):
z = -30 + 10 * i
self.obstacles.append(Obstacle(radius=5, position=[50, 0, z]))
init_pos = [0, 0, 0]
init_attitude = np.array([
0,
self.path.get_direction_angles(0)[1],
self.path.get_direction_angles(0)[0]
])
initial_state = np.hstack([init_pos, init_attitude])
return initial_state
def scenario_beginner(self):
initial_state = np.zeros(6)
self.current = Current(mu=0,
Vmin=0,
Vmax=0,
Vc_init=0,
alpha_init=0,
beta_init=0,
t_step=0) #Current object with zero velocity
waypoints = generate_random_waypoints(self.n_waypoints)
self.path = QPMI(waypoints)
init_pos = [
np.random.uniform(0, 2) * (-5),
np.random.normal(0, 1) * 5,
np.random.normal(0, 1) * 5
]
init_attitude = np.array([
0,
self.path.get_direction_angles(0)[1],
self.path.get_direction_angles(0)[0]
])
initial_state = np.hstack([init_pos, init_attitude])
return initial_state
def scenario_expert(self):
initial_state = self.scenario_advanced()
self.current = Current(mu=0.2, Vmin=0.5, Vmax=1.0, Vc_init=np.random.uniform(0.5, 1), \
alpha_init=np.random.uniform(-np.pi, np.pi), beta_init=np.random.uniform(-np.pi/4, np.pi/4), t_step=self.step_size)
self.penalize_control = 1.0
return initial_state