def reset(self):
"""
Description
-----
Resets the PID input trajectory.
"""
self.u = []
if __name__ == "__main__":
# Build RL Objects
rl = ReinforceLearning(discount_factor=0.95, states_start=300, states_stop=340, states_interval=0.5,
actions_start=-15, actions_stop=15, actions_interval=2.5, learning_rate=0.5,
epsilon=0.2, doe=1.2, eval_period=30)
# Building states for the problem, states will be the tracking errors
states = []
rl.x1 = np.zeros(20)
rl.x1[0:16] = np.linspace(-20, 0, 16)
rl.x1[16:20] = np.linspace(2, 12, 4)
rl.x2 = np.zeros(20)
rl.x2[0:3] = np.linspace(-5, 5, 3)
rl.x2[3:20] = np.linspace(6, 28, 17)
for x1 in rl.x1:
for x2 in rl.x2:
def reset(self):
"""
Description
-----
Resets the PID input trajectory.
"""
self.u = []
if __name__ == "__main__":
# Build RL Objects
rl = ReinforceLearning(discount_factor=0.95, states_start=300, states_stop=340, states_interval=0.5,
actions_start=-15, actions_stop=15, actions_interval=2.5, learning_rate=0.5,
epsilon=0.2, doe=1.2, eval_period=15, beta=0.04)
# Building states for the problem, states will be the tracking errors
states = []
rl.x1 = np.zeros(41)
rl.x1[0:41] = np.linspace(-20, 20, 41)
# rl.x2 = np.zeros(28)
# rl.x2[0:28] = np.linspace(-20, 20, 28)
rl.x2 = np.zeros(1)
for x1 in rl.x1:
for x2 in rl.x2: