def __init__(self, x=60, y=60, orientation=0, linear=1, angular=0, agent_color='r'):
self.state = State(x, y, orientation, linear, angular)
self.map = Map()
self.pixels_per_meter = 20 # TODO: Make mutator or include in constructor
self.type = agent_color
self.sensor = Sensor()
self.controller = Controller()
self.sensor_noise = NormalDistribution()
self.actuator_noise = NormalDistribution()
def __init__(self):
self.map = Map()
plt.ion()
self.figure = plt.figure(num='Robot simulator',
figsize=(8, 8),
dpi=80,
facecolor='w',
edgecolor='k')
self.figure.canvas.draw()
self.agent_posteriors = []
class Environment(object):
def __init__(self):
self.map = Map()
self.agents = []
self.visualiser = Visualiser()
self.visualise = True
self.stepsize = 0.1 # TODO: magic number...
def setSensorNoise(self, mean, variance):
self.sensor_noise.mean = mean
self.sensor_noise.variance = variance
def setActuatorNoise(self, mean, variance):
self.actuator_noise.mean = mean
self.actuator_noise.variance = variance
def takeStep(self):
for agent_index in range(len(self.agents)):
self.agents[agent_index].update()
if not self.map.inCollision((self.agents[agent_index].state.position.x, self.agents[agent_index].state.position.y)) :
if not self.agentCollision(agent_index) :
self.agents[agent_index].state.position.x = self.agents[agent_index].state.position.x + ( self.agents[agent_index].state.velocity.linear * np.cos(self.agents[agent_index].state.orientation) * self.stepsize * self.agents[agent_index].pixels_per_meter)
self.agents[agent_index].state.position.y = self.agents[agent_index].state.position.y + ( self.agents[agent_index].state.velocity.linear * np.sin(self.agents[agent_index].state.orientation) * self.stepsize * self.agents[agent_index].pixels_per_meter)
self.agents[agent_index].state.orientation = self.agents[agent_index].state.orientation + (self.agents[agent_index].state.velocity.angular * self.stepsize)
if self.visualise :
self.visualiser.plotAgent(agent_index, self.agents[agent_index])
self.visualiser.draw()
def run(self, max_iterations):
iterations = 0
while iterations < max_iterations :
iterations += 1
self.takeStep()
self.visualiser.figure.savefig('last_run.png')
print 'Completed ' + str(iterations) + ' iterations'
def agentCollision(self, index):
out = False
for agent_index in range(len(self.agents)):
if agent_index != index:
if np.abs(self.agents[agent_index].state.position.x - self.agents[index].state.position.x) < 5 and np.abs(self.agents[agent_index].state.position.y - self.agents[index].state.position.x) < 5 :
out = True
return out
def __init__(self):
self.map = Map()
self.agents = []
self.visualiser = Visualiser()
self.visualise = True
self.stepsize = 0.1 # TODO: magic number...
def __init__(self):
self.map = Map()
self.agents = []
self.visualiser = Visualiser()
self.visualise = True
self.stepsize = 0.1 # TODO: magic number...
class Environment(object):
def __init__(self):
self.map = Map()
self.agents = []
self.visualiser = Visualiser()
self.visualise = True
self.stepsize = 0.1 # TODO: magic number...
def setSensorNoise(self, mean, variance):
self.sensor_noise.mean = mean
self.sensor_noise.variance = variance
def setActuatorNoise(self, mean, variance):
self.actuator_noise.mean = mean
self.actuator_noise.variance = variance
def takeStep(self):
for agent_index in range(len(self.agents)):
self.agents[agent_index].update()
if not self.map.inCollision(
(self.agents[agent_index].state.position.x,
self.agents[agent_index].state.position.y)):
if not self.agentCollision(agent_index):
self.agents[agent_index].state.position.x = self.agents[
agent_index].state.position.x + (
self.agents[agent_index].state.velocity.linear *
np.cos(self.agents[agent_index].state.orientation)
* self.stepsize *
self.agents[agent_index].pixels_per_meter)
self.agents[agent_index].state.position.y = self.agents[
agent_index].state.position.y + (
self.agents[agent_index].state.velocity.linear *
np.sin(self.agents[agent_index].state.orientation)
* self.stepsize *
self.agents[agent_index].pixels_per_meter)
self.agents[agent_index].state.orientation = self.agents[
agent_index].state.orientation + (
self.agents[agent_index].state.velocity.angular *
self.stepsize)
if self.visualise:
self.visualiser.plotAgent(agent_index,
self.agents[agent_index])
self.visualiser.draw()
def run(self, max_iterations):
iterations = 0
while iterations < max_iterations:
iterations += 1
self.takeStep()
self.visualiser.figure.savefig('last_run.png')
print 'Completed ' + str(iterations) + ' iterations'
def agentCollision(self, index):
out = False
for agent_index in range(len(self.agents)):
if agent_index != index:
if np.abs(self.agents[agent_index].state.position.x -
self.agents[index].state.position.x) < 5 and np.abs(
self.agents[agent_index].state.position.y -
self.agents[index].state.position.x) < 5:
out = True
return out
def __init__(self):
self.map = Map()
