def __init__(self,
name,
vehicles,
net_params,
initial_config=InitialConfig(),
traffic_lights=TrafficLightParams(),
detector_params=None):
"""Instantiate the base network class.
Attributes
----------
name : str
A tag associated with the network
vehicles : flow.core.params.VehicleParams
see flow/core/params.py
net_params : flow.core.params.NetParams
see flow/core/params.py
initial_config : flow.core.params.InitialConfig
see flow/core/params.py
traffic_lights : flow.core.params.TrafficLightParams
see flow/core/params.py
"""
self.orig_name = name # To avoid repeated concatenation upon reset
self.name = name + time.strftime('_%Y%m%d-%H%M%S') + str(time.time())
self.vehicles = vehicles
self.net_params = net_params
self.initial_config = initial_config
self.traffic_lights = traffic_lights
self.detector_params = detector_params if detector_params else DetectorParams(
)
# specify routes vehicles can take
self.routes = self.specify_routes(net_params)
if net_params.template is None and net_params.osm_path is None:
# specify the attributes of the nodes
self.nodes = self.specify_nodes(net_params)
# collect the attributes of each edge
self.edges = self.specify_edges(net_params)
# specify the types attributes (default is None)
self.types = self.specify_types(net_params)
# specify the connection attributes (default is None)
self.connections = self.specify_connections(net_params)
# this is to be used if file paths other than the the network geometry
# file is specified
elif type(net_params.template) is dict:
if 'rou' in net_params.template:
veh, rou = self._vehicle_infos(net_params.template['rou'])
vtypes = self._vehicle_type(net_params.template.get('vtype'))
cf = self._get_cf_params(vtypes)
lc = self._get_lc_params(vtypes)
# add the vehicle types to the VehicleParams object
for t in vtypes:
vehicles.add(veh_id=t,
length=vtypes[t]['length'],
car_following_params=cf[t],
lane_change_params=lc[t],
num_vehicles=0)
# add the routes of the vehicles that will be departed later
# under the name of the vehicle. This will later be identified
# by k.vehicles._add_departed
self.routes = rou
# vehicles to be added with different departure times
self.template_vehicles = veh
if 'add' in net_params.template:
e1Detectors, e2Detectors = self._get_detector_params(
net_params.template['add'])
for e1Detector in e1Detectors:
self.detector_params.add_induction_loop_detector(
name=e1Detector['id'],
lane_id=e1Detector['lane'],
position=e1Detector['pos'],
frequency=e1Detector['freq'],
storage_file=e1Detector['file'],
)
for e2Detector in e2Detectors:
self.detector_params.add_lane_area_detector(
name=e2Detector['id'],
lane_id=e2Detector['lane'],
position=e2Detector['pos'],
frequency=e2Detector['freq'],
storage_file=e2Detector['file'],
)
self.types = None
self.nodes = None
self.edges = None
self.connections = None
# osm_path or template as type str
else:
self.nodes = None
self.edges = None
self.types = None
self.connections = None
# optional parameters, used to get positions from some global reference
self.edge_starts = self.specify_edge_starts()
self.internal_edge_starts = self.specify_internal_edge_starts()
self.intersection_edge_starts = [] # this will be deprecated
def figure_eight_baseline(num_runs, render=True):
"""Run script for all figure eight baselines.
Parameters
----------
num_runs : int
number of rollouts the performance of the environment is evaluated
over
render : bool, optional
specifies whether to use the gui during execution
Returns
-------
Experiment
class needed to run simulations
"""
exp_tag = flow_params['exp_tag']
sim_params = flow_params['sim']
env_params = flow_params['env']
net_params = flow_params['net']
initial_config = flow_params.get('initial', InitialConfig())
traffic_lights = flow_params.get('tls', TrafficLightParams())
# modify the rendering to match what is requested
sim_params.render = render
# set the evaluation flag to True
env_params.evaluate = True
# we want no autonomous vehicles in the simulation
vehicles = VehicleParams()
vehicles.add(veh_id='human',
acceleration_controller=(IDMController, {
'noise': 0.2
}),
routing_controller=(ContinuousRouter, {}),
car_following_params=SumoCarFollowingParams(
speed_mode='no_collide', ),
num_vehicles=14)
# import the scenario class
module = __import__('flow.scenarios', fromlist=[flow_params['scenario']])
scenario_class = getattr(module, flow_params['scenario'])
# create the scenario object
scenario = scenario_class(name=exp_tag,
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights)
# import the environment class
module = __import__('flow.envs', fromlist=[flow_params['env_name']])
env_class = getattr(module, flow_params['env_name'])
# create the environment object
env = env_class(env_params, sim_params, scenario)
exp = Experiment(env)
results = exp.run(num_runs, env_params.horizon)
avg_speed = np.mean(results['mean_returns'])
return avg_speed
def bottleneck2_baseline(num_runs, render=True):
"""Run script for the bottleneck2 baseline.
Parameters
----------
num_runs : int
number of rollouts the performance of the environment is evaluated
over
render : bool, optional
specifies whether to use sumo's gui during execution
Returns
-------
flow.core.experiment.Experiment
class needed to run simulations
"""
exp_tag = flow_params['exp_tag']
sumo_params = flow_params['sumo']
env_params = flow_params['env']
net_params = flow_params['net']
initial_config = flow_params.get('initial', InitialConfig())
traffic_lights = flow_params.get('tls', TrafficLightParams())
# we want no autonomous vehicles in the simulation
vehicles = Vehicles()
vehicles.add(veh_id='human',
sumo_car_following_params=SumoCarFollowingParams(
speed_mode=9,
),
routing_controller=(ContinuousRouter, {}),
sumo_lc_params=SumoLaneChangeParams(
lane_change_mode=0,
),
num_vehicles=1 * SCALING)
# only include human vehicles in inflows
flow_rate = 1900 * SCALING
inflow = InFlows()
inflow.add(veh_type='human', edge='1',
vehs_per_hour=flow_rate,
departLane='random', departSpeed=10)
net_params.inflows = inflow
# modify the rendering to match what is requested
sumo_params.render = render
# set the evaluation flag to True
env_params.evaluate = True
# import the scenario class
module = __import__('flow.scenarios', fromlist=[flow_params['scenario']])
scenario_class = getattr(module, flow_params['scenario'])
# create the scenario object
scenario = scenario_class(
name=exp_tag,
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights
)
# import the environment class
module = __import__('flow.envs', fromlist=[flow_params['env_name']])
env_class = getattr(module, flow_params['env_name'])
# create the environment object
env = env_class(env_params, sumo_params, scenario)
exp = Experiment(env)
results = exp.run(num_runs, env_params.horizon)
return np.mean(results['returns']), np.std(results['returns'])
def traffic_light_grid_example(render=None):
"""
Perform a simulation of vehicles on a traffic light grid.
Parameters
----------
render: bool, optional
specifies whether to use the gui during execution
Returns
-------
exp: flow.core.experiment.Experiment
A non-rl experiment demonstrating the performance of human-driven
vehicles and balanced traffic lights on a traffic light grid.
"""
inner_length = 300
long_length = 500
short_length = 300
N_ROWS = 2
N_COLUMNS = 3
num_cars_left = 20
num_cars_right = 20
num_cars_top = 20
num_cars_bot = 20
tot_cars = (num_cars_left + num_cars_right) * N_COLUMNS \
+ (num_cars_top + num_cars_bot) * N_ROWS
grid_array = {
"short_length": short_length,
"inner_length": inner_length,
"long_length": long_length,
"row_num": N_ROWS,
"col_num": N_COLUMNS,
"cars_left": num_cars_left,
"cars_right": num_cars_right,
"cars_top": num_cars_top,
"cars_bot": num_cars_bot
}
sim_params = AimsunParams(sim_step=0.5, render=True)
if render is not None:
sim_params.render = render
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
num_vehicles=tot_cars)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"yellow": "3",
"state": "GGGrrrGGGrrr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"yellow": "3",
"state": "yyyrrryyyrrr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"yellow": "3",
"state": "rrrGGGrrrGGG"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"yellow": "3",
"state": "rrryyyrrryyy"
}]
tl_logic.add("center0", phases=phases, programID=1)
tl_logic.add("center1", phases=phases, programID=1)
tl_logic.add("center2", tls_type="actuated", phases=phases, programID=1)
tl_logic.add("center3", phases=phases, programID=1)
tl_logic.add("center4", phases=phases, programID=1)
tl_logic.add("center5", tls_type="actuated", phases=phases, programID=1)
additional_net_params = {
"grid_array": grid_array,
"speed_limit": 35,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
net_params = NetParams(additional_params=additional_net_params)
initial_config = InitialConfig(spacing='custom')
network = TrafficLightGridNetwork(
name="grid-intersection",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
env = AccelEnv(env_params, sim_params, network, simulator='aimsun')
return Experiment(env)
def grid_mxn_exp_setup(row_num=1,
col_num=1,
sim_params=None,
vehicles=None,
env_params=None,
net_params=None,
initial_config=None,
tl_logic=None):
"""
Create an environment and scenario pair for grid 1x1 test experiments.
Parameters
----------
row_num: int, optional
number of horizontal rows of edges in the grid network
col_num: int, optional
number of vertical columns of edges in the grid network
sim_params : flow.core.params.SumoParams
sumo-related configuration parameters, defaults to a time step of 1s
and no sumo-imposed failsafe on human or rl vehicles
vehicles : Vehicles type
vehicles to be placed in the network, default is 5 vehicles per edge
for a total of 20 vehicles with an IDM acceleration controller and
GridRouter routing controller.
env_params : flow.core.params.EnvParams
environment-specific parameters, defaults to a environment with
failsafes, where other parameters do not matter for non-rl runs
net_params : flow.core.params.NetParams
network-specific configuration parameters, defaults to a 1x1 grid
which traffic lights on and "no_internal_links" set to False
initial_config : flow.core.params.InitialConfig
specifies starting positions of vehicles, defaults to evenly
distributed vehicles across the length of the network
tl_logic: flow.core.params.TrafficLightParams
specifies logic of any traffic lights added to the system
"""
logging.basicConfig(level=logging.WARNING)
if tl_logic is None:
tl_logic = TrafficLightParams(baseline=False)
if sim_params is None:
# set default sim_params configuration
sim_params = SumoParams(sim_step=1, render=False)
if vehicles is None:
total_vehicles = 20
vehicles = VehicleParams()
vehicles.add(veh_id="idm",
acceleration_controller=(IDMController, {}),
car_following_params=SumoCarFollowingParams(min_gap=2.5,
tau=1.1,
max_speed=30),
routing_controller=(GridRouter, {}),
num_vehicles=total_vehicles)
if env_params is None:
# set default env_params configuration
additional_env_params = {
"target_velocity": 50,
"switch_time": 3.0,
"tl_type": "controlled",
"discrete": False
}
env_params = EnvParams(additional_params=additional_env_params,
horizon=100)
if net_params is None:
# set default net_params configuration
total_vehicles = vehicles.num_vehicles
num_entries = 2 * row_num + 2 * col_num
grid_array = {
"short_length": 100,
"inner_length": 300,
"long_length": 3000,
"row_num": row_num,
"col_num": col_num,
"cars_left": int(total_vehicles / num_entries),
"cars_right": int(total_vehicles / num_entries),
"cars_top": int(total_vehicles / num_entries),
"cars_bot": int(total_vehicles / num_entries)
}
additional_net_params = {
"length": 200,
"lanes": 2,
"speed_limit": 35,
"resolution": 40,
"grid_array": grid_array,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
net_params = NetParams(no_internal_links=False,
additional_params=additional_net_params)
if initial_config is None:
# set default initial_config configuration
initial_config = InitialConfig(spacing="custom",
additional_params={"enter_speed": 30})
# create the scenario
scenario = SimpleGridScenario(name="Grid1x1Test",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
# create the environment
env = GreenWaveTestEnv(env_params=env_params,
sim_params=sim_params,
scenario=scenario)
# reset the environment
env.reset()
return env, scenario
def traffic_light_grid_mxn_exp_setup(row_num=1,
col_num=1,
sim_params=None,
vehicles=None,
env_params=None,
net_params=None,
initial_config=None,
tl_logic=None):
"""
Create an environment and network pair for traffic light grid 1x1 test experiments.
Parameters
----------
row_num: int, optional
number of horizontal rows of edges in the traffic light grid network
col_num: int, optional
number of vertical columns of edges in the traffic light grid network
sim_params : flow.core.params.SumoParams
sumo-related configuration parameters, defaults to a time step of 1s
and no sumo-imposed failsafe on human or rl vehicles
vehicles : Vehicles type
vehicles to be placed in the network, default is 5 vehicles per edge
for a total of 20 vehicles with an IDM acceleration controller and
GridRouter routing controller.
env_params : flow.core.params.EnvParams
environment-specific parameters, defaults to a environment with
failsafes, where other parameters do not matter for non-rl runs
net_params : flow.core.params.NetParams
network-specific configuration parameters, defaults to a 1x1 traffic
light grid with traffic lights on
initial_config : flow.core.params.InitialConfig
specifies starting positions of vehicles, defaults to evenly
distributed vehicles across the length of the network
tl_logic: flow.core.params.TrafficLightParams
specifies logic of any traffic lights added to the system
"""
logging.basicConfig(level=logging.WARNING)
if tl_logic is None:
tl_logic = TrafficLightParams(baseline=False)
if sim_params is None:
# set default sim_params configuration
sim_params = SumoParams(sim_step=1, render=False)
if vehicles is None:
vehicles_per_edge = 5
num_edges = 2 * (row_num + col_num)
total_vehicles = num_edges * vehicles_per_edge
vehicles = VehicleParams()
vehicles.add(veh_id="idm",
acceleration_controller=(IDMController, {}),
car_following_params=SumoCarFollowingParams(min_gap=2.5,
tau=1.1,
max_speed=30),
routing_controller=(GridRouter, {}),
num_vehicles=total_vehicles)
if env_params is None:
# set default env_params configuration
additional_env_params = {
"target_velocity": 50,
"switch_time": 3.0,
"tl_type": "controlled",
"discrete": False
}
env_params = EnvParams(additional_params=additional_env_params,
horizon=100)
if net_params is None:
# set default net_params configuration
total_vehicles = vehicles.num_vehicles
num_entries = 2 * row_num + 2 * col_num
assert total_vehicles % num_entries == 0, "{} total vehicles should " \
"be divisible by {" \
"}".format(total_vehicles,
num_entries)
grid_array = {
"short_length": 100,
"inner_length": 300,
"long_length": 3000,
"row_num": row_num,
"col_num": col_num,
"cars_left": int(total_vehicles / num_entries),
"cars_right": int(total_vehicles / num_entries),
"cars_top": int(total_vehicles / num_entries),
"cars_bot": int(total_vehicles / num_entries)
}
additional_net_params = {
"length": 200,
"lanes": 2,
"speed_limit": 35,
"resolution": 40,
"grid_array": grid_array,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
net_params = NetParams(additional_params=additional_net_params)
if initial_config is None:
# set default initial_config configuration
initial_config = InitialConfig(spacing="custom",
additional_params={"enter_speed": 30})
flow_params = dict(
# name of the experiment
exp_tag="Grid1x1Test",
# name of the flow environment the experiment is running on
env_name=TrafficLightGridTestEnv,
# name of the network class the experiment is running on
network=TrafficLightGridNetwork,
# simulator that is used by the experiment
simulator='traci',
# sumo-related parameters (see flow.core.params.SumoParams)
sim=sim_params,
# environment related parameters (see flow.core.params.EnvParams)
env=env_params,
# network-related parameters (see flow.core.params.NetParams and the
# network's documentation or ADDITIONAL_NET_PARAMS component)
net=net_params,
# vehicles to be placed in the network at the start of a rollout (see
# flow.core.params.VehicleParams)
veh=vehicles,
# parameters specifying the positioning of vehicles upon initialization/
# reset (see flow.core.params.InitialConfig)
initial=initial_config,
# traffic lights to be introduced to specific nodes (see
# flow.core.params.TrafficLightParams)
tls=tl_logic)
# create the network
network = TrafficLightGridNetwork(name="Grid1x1Test",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
# create the environment
env = TrafficLightGridTestEnv(env_params=env_params,
sim_params=sim_params,
network=network)
# reset the environment
env.reset()
return env, network, flow_params
def __init__(self,
name,
vehicles,
net_params,
initial_config=InitialConfig(),
traffic_lights=TrafficLightParams()):
"""Initialize an nxm grid scenario.
The grid scenario consists of m vertical lanes and n horizontal lanes,
with a total of nxm intersections where the vertical and horizontal
edges meet.
Requires from net_params:
- grid_array: dictionary of grid array data, with the following keys
- row_num: number of horizontal rows of edges
- col_num: number of vertical columns of edges
- inner_length: length of inner edges in the grid network
- short_length: length of edges that vehicles start on
- long_length: length of final edge in route
- cars_top: number of cars starting at the edges heading to the top
- cars_bot: number of cars starting at the edges heading to the
bottom
- cars_left: number of cars starting at the edges heading to the left
- cars_right: number of cars starting at the edges heading to the
right
- horizontal_lanes: number of lanes in the horizontal edges
- vertical_lanes: number of lanes in the vertical edges
- speed_limit: speed limit for all edges. This may be represented as a
float value, or a dictionary with separate values for vertical and
horizontal lanes.
In order for right-of-way dynamics to take place at the intersections,
set "no_internal_links" in net_params to False.
See flow/scenarios/base_scenario.py for description of params.
"""
optional = ["tl_logic"]
for p in ADDITIONAL_NET_PARAMS.keys():
if p not in net_params.additional_params and p not in optional:
raise KeyError('Network parameter "{}" not supplied'.format(p))
for p in ADDITIONAL_NET_PARAMS["grid_array"].keys():
if p not in net_params.additional_params["grid_array"]:
raise KeyError(
'Grid array parameter "{}" not supplied'.format(p))
# this is a (mx1)x(nx1)x2 array
# the third dimension is vertical length, horizontal length
self.grid_array = net_params.additional_params["grid_array"]
vertical_lanes = net_params.additional_params["vertical_lanes"]
horizontal_lanes = net_params.additional_params["horizontal_lanes"]
self.horizontal_junction_len = 2.9 + 3.3 * vertical_lanes
self.vertical_junction_len = 2.9 + 3.3 * horizontal_lanes
self.row_num = self.grid_array["row_num"]
self.col_num = self.grid_array["col_num"]
self.num_edges = (self.col_num+1) * self.row_num * 2 \
+ (self.row_num+1) * self.col_num * 2 + self.row_num * self.col_num
self.inner_length = self.grid_array["inner_length"]
self.short_length = self.grid_array["short_length"]
self.long_length = self.grid_array["long_length"]
# this is a dictionary containing inner length, long outer length,
# short outer length, and number of rows and columns
self.grid_array = net_params.additional_params["grid_array"]
self.node_mapping = defaultdict(list)
self.name = "BobLoblawsLawBlog" # DO NOT CHANGE
super().__init__(name, vehicles, net_params, initial_config,
traffic_lights)
def setup_flow_params(args):
DISABLE_TB = True
DISABLE_RAMP_METER = True
av_frac = args.av_frac
if args.lc_on:
lc_mode = 1621
else:
lc_mode = 0
vehicles = VehicleParams()
if not np.isclose(av_frac, 1):
vehicles.add(veh_id="human",
lane_change_controller=(SimLaneChangeController, {}),
routing_controller=(ContinuousRouter, {}),
car_following_params=SumoCarFollowingParams(
speed_mode=31, ),
lane_change_params=SumoLaneChangeParams(
lane_change_mode=lc_mode, ),
num_vehicles=1)
vehicles.add(
veh_id="av",
acceleration_controller=(RLController, {}),
lane_change_controller=(SimLaneChangeController, {}),
routing_controller=(ContinuousRouter, {}),
car_following_params=SumoCarFollowingParams(speed_mode=31, ),
lane_change_params=SumoLaneChangeParams(lane_change_mode=0, ),
num_vehicles=1)
else:
vehicles.add(
veh_id="av",
acceleration_controller=(RLController, {}),
lane_change_controller=(SimLaneChangeController, {}),
routing_controller=(ContinuousRouter, {}),
car_following_params=SumoCarFollowingParams(speed_mode=31, ),
lane_change_params=SumoLaneChangeParams(lane_change_mode=0, ),
num_vehicles=1)
# flow rate
flow_rate = 1900 * args.scaling
controlled_segments = [('1', 1, False), ('2', 2, True), ('3', 2, True),
('4', 2, True), ('5', 1, False)]
num_observed_segments = [('1', 1), ('2', 3), ('3', 3), ('4', 3), ('5', 1)]
if np.isclose(args.av_frac, 0.4):
q_init = 1000
else:
q_init = 600
additional_env_params = {
'target_velocity': 40,
'disable_tb': True,
'disable_ramp_metering': True,
'controlled_segments': controlled_segments,
'symmetric': False,
'observed_segments': num_observed_segments,
'reset_inflow': True,
'lane_change_duration': 5,
'max_accel': 3,
'max_decel': 3,
'inflow_range': [args.low_inflow, args.high_inflow],
'start_inflow': flow_rate,
'congest_penalty': args.congest_penalty,
'communicate': args.communicate,
"centralized_obs": args.central_obs,
"aggregate_info": args.aggregate_info,
"av_frac": args.av_frac,
"congest_penalty_start": args.congest_penalty_start,
"lc_mode": lc_mode,
"life_penalty": args.life_penalty,
'keep_past_actions': args.keep_past_actions,
"num_sample_seconds": args.num_sample_seconds,
"speed_reward": args.speed_reward,
'fair_reward': False, # This doesn't do anything, remove
'exit_history_seconds': 0, # This doesn't do anything, remove
# parameters for the staggering controller that we imitate
"n_crit": 8,
"q_max": 15000,
"q_min": 200,
"q_init": q_init, #
"feedback_coeff": 1, #
'num_imitation_iters': args.num_imitation_iters,
}
# percentage of flow coming out of each lane
inflow = InFlows()
if not np.isclose(args.av_frac, 1.0):
inflow.add(veh_type='human',
edge='1',
vehs_per_hour=flow_rate * (1 - args.av_frac),
departLane='random',
departSpeed=23.0)
inflow.add(veh_type='av',
edge='1',
vehs_per_hour=flow_rate * args.av_frac,
departLane='random',
departSpeed=23.0)
else:
inflow.add(veh_type='av',
edge='1',
vehs_per_hour=flow_rate,
departLane='random',
departSpeed=23.0)
traffic_lights = TrafficLightParams()
if not DISABLE_TB:
traffic_lights.add(node_id='2')
if not DISABLE_RAMP_METER:
traffic_lights.add(node_id='3')
additional_net_params = {'scaling': args.scaling, "speed_limit": 23.0}
if args.imitate:
env_name = 'MultiBottleneckImitationEnv'
else:
env_name = 'MultiBottleneckEnv'
flow_params = dict(
# name of the experiment
exp_tag=args.exp_title,
# name of the flow environment the experiment is running on
env_name=env_name,
# name of the scenario class the experiment is running on
scenario='BottleneckScenario',
# simulator that is used by the experiment
simulator='traci',
# sumo-related parameters (see flow.core.params.SumoParams)
sim=SumoParams(
sim_step=args.sim_step,
render=args.render,
print_warnings=False,
restart_instance=True,
),
# environment related parameters (see flow.core.params.EnvParams)
env=EnvParams(
warmup_steps=int(0 / args.sim_step),
sims_per_step=2,
horizon=args.horizon,
clip_actions=False,
additional_params=additional_env_params,
),
# network-related parameters (see flow.core.params.NetParams and the
# scenario's documentation or ADDITIONAL_NET_PARAMS component)
net=NetParams(
inflows=inflow,
no_internal_links=False,
additional_params=additional_net_params,
),
# vehicles to be placed in the network at the start of a rollout (see
# flow.core.vehicles.Vehicles)
veh=vehicles,
# parameters specifying the positioning of vehicles upon initialization/
# reset (see flow.core.params.InitialConfig)
initial=InitialConfig(
spacing='uniform',
min_gap=5,
lanes_distribution=float('inf'),
edges_distribution=['2', '3', '4', '5'],
),
# traffic lights to be introduced to specific nodes (see
# flow.core.traffic_lights.TrafficLights)
tls=traffic_lights,
)
return flow_params
from flow.core.params import SumoParams # de parameters die we nog is hebben om sumo te runnen
from flow.core.params import EnvParams
from flow.core.experiment import Experiment #opzetten van het expirement
# ?
from flow.controllers.car_following_models import IDMController
# reroute all vehicles to the initial set route
from flow.controllers.routing_controllers import ContinuousRouter
# toevoegen van extra parameters
from flow.networks.ring import ADDITIONAL_NET_PARAMS
from flow.envs.ring.accel import ADDITIONAL_ENV_PARAMS
# naam van de simulatie
name = "ring_example"
#creating empty vehicleParamsobject/ trafficlightobject
vehicles = VehicleParams()
trafficL = TrafficLightParams()
# simulation parameters
sim_params = SumoParams(sim_step=0.1, render=True, emission_path='data')
# 22 voertuigen toevoegen met bovenstaande acceleration and routing behavior
vehicles.add("human",
acceleration_controller=(IDMController, {}),
routing_controller=(ContinuousRouter, {}),
num_vehicles=22)
net_params = NetParams(additional_params=ADDITIONAL_NET_PARAMS)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
# initial config voor de simulatie,
initial_config = InitialConfig(spacing="uniform", perturbation=1)
# vinden we de nodigeparameters mee terug
print(ADDITIONAL_NET_PARAMS)
print("runnen v/ h expirement")
def run_task(*_):
"""Implement the run_task method needed to run experiments with rllab."""
V_ENTER = 30
INNER_LENGTH = 300
LONG_LENGTH = 100
SHORT_LENGTH = 300
N_ROWS = 3
N_COLUMNS = 3
NUM_CARS_LEFT = 1
NUM_CARS_RIGHT = 1
NUM_CARS_TOP = 1
NUM_CARS_BOT = 1
tot_cars = (NUM_CARS_LEFT + NUM_CARS_RIGHT) * N_COLUMNS \
+ (NUM_CARS_BOT + NUM_CARS_TOP) * N_ROWS
grid_array = {
"short_length": SHORT_LENGTH,
"inner_length": INNER_LENGTH,
"long_length": LONG_LENGTH,
"row_num": N_ROWS,
"col_num": N_COLUMNS,
"cars_left": NUM_CARS_LEFT,
"cars_right": NUM_CARS_RIGHT,
"cars_top": NUM_CARS_TOP,
"cars_bot": NUM_CARS_BOT
}
sim_params = SumoParams(sim_step=1, render=True)
vehicles = VehicleParams()
vehicles.add(veh_id="idm",
acceleration_controller=(SimCarFollowingController, {}),
car_following_params=SumoCarFollowingParams(
min_gap=2.5,
tau=1.1,
max_speed=V_ENTER,
speed_mode="all_checks"),
routing_controller=(GridRouter, {}),
num_vehicles=tot_cars)
tl_logic = TrafficLightParams(baseline=False)
additional_env_params = {
"target_velocity": 50,
"switch_time": 3.0,
"num_observed": 2,
"discrete": False,
"tl_type": "controlled"
}
env_params = EnvParams(additional_params=additional_env_params)
additional_net_params = {
"speed_limit": 35,
"grid_array": grid_array,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
if USE_INFLOWS:
initial_config, net_params = get_flow_params(
v_enter=V_ENTER,
vehs_per_hour=EDGE_INFLOW,
col_num=N_COLUMNS,
row_num=N_ROWS,
add_net_params=additional_net_params)
else:
initial_config, net_params = get_non_flow_params(
V_ENTER, additional_net_params)
scenario = SimpleGridScenario(name="grid-intersection",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
env_name = "PO_TrafficLightGridEnv"
pass_params = (env_name, sim_params, vehicles, env_params, net_params,
initial_config, scenario)
env = GymEnv(env_name, record_video=False, register_params=pass_params)
horizon = env.horizon
env = normalize(env)
policy = GaussianMLPPolicy(env_spec=env.spec, hidden_sizes=(32, 32))
baseline = LinearFeatureBaseline(env_spec=env.spec)
algo = TRPO(
env=env,
policy=policy,
baseline=baseline,
batch_size=40000,
max_path_length=horizon,
# whole_paths=True,
n_itr=800,
discount=0.999,
# step_size=0.01,
)
algo.train()
def grid1_baseline(num_runs, render=True):
"""Run script for the grid1 baseline.
Parameters
----------
num_runs : int
number of rollouts the performance of the environment is evaluated
over
render: bool, optional
specifies whether to the gui during execution
Returns
-------
flow.core.experiment.Experiment
class needed to run simulations
"""
sim_params = flow_params['sim']
env_params = flow_params['env']
# define the traffic light logic
tl_logic = TrafficLightParams(baseline=False)
phases = [{
'duration': '31',
'minDur': '5',
'maxDur': '45',
"state": "GrGr"
}, {
'duration': '2',
'minDur': '2',
'maxDur': '2',
"state": "yryr"
}, {
'duration': '31',
'minDur': '5',
'maxDur': '45',
"state": "rGrG"
}, {
'duration': '2',
'minDur': '2',
'maxDur': '2',
"state": "ryry"
}]
for i in range(N_ROWS * N_COLUMNS):
tl_logic.add('center' + str(i),
tls_type='actuated',
phases=phases,
programID=1)
# modify the rendering to match what is requested
sim_params.render = render
# set the evaluation flag to True
env_params.evaluate = True
flow_params['env'].horizon = env_params.horizon
exp = Experiment(flow_params)
results = exp.run(num_runs)
total_delay = np.mean(results['returns'])
return total_delay
def traffic_light_grid_example(render=None, use_inflows=False):
"""
Perform a simulation of vehicles on a traffic light grid.
Parameters
----------
render: bool, optional
specifies whether to use the gui during execution
use_inflows : bool, optional
set to True if you would like to run the experiment with inflows of
vehicles from the edges, and False otherwise
Returns
-------
exp: flow.core.experiment.Experiment
A non-rl experiment demonstrating the performance of human-driven
vehicles and balanced traffic lights on a traffic light grid.
"""
v_enter = 10
inner_length = 300
long_length = 500
short_length = 300
n_rows = 2
n_columns = 3
num_cars_left = 20
num_cars_right = 20
num_cars_top = 20
num_cars_bot = 20
tot_cars = (num_cars_left + num_cars_right) * n_columns \
+ (num_cars_top + num_cars_bot) * n_rows
grid_array = {
"short_length": short_length,
"inner_length": inner_length,
"long_length": long_length,
"row_num": n_rows,
"col_num": n_columns,
"cars_left": num_cars_left,
"cars_right": num_cars_right,
"cars_top": num_cars_top,
"cars_bot": num_cars_bot
}
sim_params = OTMParams(sim_step=0.1, render=True)
if render is not None:
sim_params.render = render
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
routing_controller=(GridRouter, {}),
# car_following_params=SumoCarFollowingParams(
# min_gap=2.5,
# decel=7.5, # avoid collisions at emergency stops
# ),
num_vehicles=tot_cars)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GrGrGrGrGrGr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yryryryryryr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rGrGrGrGrGrG"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "ryryryryryry"
}]
tl_logic.add("center0", phases=phases, programID=1)
tl_logic.add("center1", phases=phases, programID=1)
tl_logic.add("center2", phases=phases, programID=1, tls_type="actuated")
additional_net_params = {
"grid_array": grid_array,
"speed_limit": 35,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
if use_inflows:
initial_config, net_params = get_flow_params(
col_num=n_columns,
row_num=n_rows,
additional_net_params=additional_net_params)
else:
initial_config, net_params = get_non_flow_params(
enter_speed=v_enter, add_net_params=additional_net_params)
network = TrafficLightGridNetwork(name="grid-intersection",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
env = TestEnv(env_params, sim_params, network, simulator='otm')
return Experiment(env)
# all other imports are standard
from flow.core.params import VehicleParams
from flow.core.params import NetParams
from flow.core.params import InitialConfig
from flow.core.params import EnvParams
from flow.core.params import SumoParams
from flow.core.params import TrafficLightParams
# map data
from Network.osm_network import OsmNetwork
from flow.envs.base import Env
import gym
from abc import ABCMeta
# if traffic is in osm, activate this
tl_logic = TrafficLightParams(baseline=False)
net_params = NetParams(osm_path='./Network/map.osm')
env_params = EnvParams()
sim_params = SumoParams(render=True)
initial_config = InitialConfig()
vehicles = VehicleParams()
vehicles.add(
'human',
num_vehicles=80,
)
flow_params = dict(
exp_tag='osm_test',
env_name=TestEnv,
network=OsmNetwork,
def __init__(self,
name,
vehicles,
net_params,
initial_config=InitialConfig(),
traffic_lights=TrafficLightParams()):
"""Initialize a highway with on and off ramps network."""
for p in ADDITIONAL_NET_PARAMS.keys():
if p not in net_params.additional_params:
raise KeyError('Network parameter "{}" not supplied'.format(p))
# load parameters into class
params = net_params.additional_params
self.highway_length = params['highway_length']
self.on_ramps_length = params['on_ramps_length']
self.off_ramps_length = params['off_ramps_length']
self.highway_lanes = params['highway_lanes']
self.on_ramps_lanes = params['on_ramps_lanes']
self.off_ramps_lanes = params['off_ramps_lanes']
self.highway_speed = params['highway_speed']
self.on_ramps_speed = params['on_ramps_speed']
self.off_ramps_speed = params['off_ramps_speed']
self.on_ramps_pos = params['on_ramps_pos']
self.off_ramps_pos = params['off_ramps_pos']
self.p = params['next_off_ramp_proba']
self.angle_on_ramps = params['angle_on_ramps']
self.angle_off_ramps = params['angle_off_ramps']
# generate position of all network nodes
self.ramps_pos = sorted(self.on_ramps_pos + self.off_ramps_pos)
self.nodes_pos = sorted(
list(set([0] + self.ramps_pos + [self.highway_length])))
# highway_pos[x] = id of the highway node whose starting position is x
self.highway_pos = {x: i for i, x in enumerate(self.nodes_pos)}
# ramp_pos[x] = id of the ramp node whose intersection with the highway
# is at position x
self.ramp_pos = {
x: "on_ramp_{}".format(i)
for i, x in enumerate(self.on_ramps_pos)
}
self.ramp_pos.update({
x: "off_ramp_{}".format(i)
for i, x in enumerate(self.off_ramps_pos)
})
# make sure network is constructable
if (len(self.ramps_pos) > 0
and (min(self.ramps_pos) <= 0
or max(self.ramps_pos) >= self.highway_length)):
raise ValueError('All ramps positions should be positive and less '
'than highway length. Current ramps positions: {}'
'. Current highway length: {}.'.format(
self.ramps_pos, self.highway_length))
if len(self.ramps_pos) != len(list(set(self.ramps_pos))):
raise ValueError('Two ramps positions cannot be equal.')
super().__init__(name, vehicles, net_params, initial_config,
traffic_lights)
# features associated with the routes vehicles take
"vtype":
"/home/rong/Safe-RL-for-Driving/traci_pedestrian_crossing/pedcrossing.add.xml",
# 和下方specify_routes一致
"rou":
"/home/rong/Safe-RL-for-Driving/traci_pedestrian_crossing/data/pedcrossing.rou.xml",
"trip":
"/home/rong/Safe-RL-for-Driving/traci_pedestrian_crossing/pedestrians.trip.xml"
},
inflows=inflow,
)
# specify the edges vehicles can originate on
initial_config = InitialConfig(edges_distribution=["WC"])
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "100000",
"state": "GGGGr"
}, {
"duration": "4",
"state": "yyyyr"
}, {
"duration": "10",
"state": "rrrrG"
}, {
"duration": "10",
"state": "rrrrr"
}]
tl_logic.add("C", phases=phases, programID="custom", offset="0")
def evaluate_policy(benchmark, _get_actions, _get_states=None):
"""Evaluate the performance of a controller on a predefined benchmark.
Parameters
----------
benchmark : str
name of the benchmark, must be printed as it is in the
benchmarks folder; otherwise a ValueError will be raised
_get_actions : method
the mapping from states to actions for the RL agent(s)
_get_states : method, optional
a mapping from the environment object in Flow to some state, which
overrides the _get_states method of the environment. Note that the
same cannot be done for the actions.
Returns
-------
float
mean of the evaluation return of the benchmark from NUM_RUNS number
of simulations
float
standard deviation of the evaluation return of the benchmark from
NUM_RUNS number of simulations
Raises
------
ValueError
If the specified benchmark is not available.
"""
if benchmark not in AVAILABLE_BENCHMARKS.keys():
raise ValueError(
"benchmark {} is not available. Check spelling?".format(benchmark))
# get the flow params from the benchmark
flow_params = AVAILABLE_BENCHMARKS[benchmark]
exp_tag = flow_params["exp_tag"]
sim_params = flow_params["sim"]
vehicles = flow_params["veh"]
env_params = flow_params["env"]
env_params.evaluate = True # Set to true to get evaluation returns
net_params = flow_params["net"]
initial_config = flow_params.get("initial", InitialConfig())
traffic_lights = flow_params.get("tls", TrafficLightParams())
# import the environment and scenario classes
module = __import__("flow.envs", fromlist=[flow_params["env_name"]])
env_class = getattr(module, flow_params["env_name"])
module = __import__("flow.scenarios", fromlist=[flow_params["scenario"]])
scenario_class = getattr(module, flow_params["scenario"])
# recreate the scenario and environment
scenario = scenario_class(
name=exp_tag,
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights)
# make sure the _get_states method of the environment is the one
# specified by the user
if _get_states is not None:
class _env_class(env_class):
def get_state(self):
return _get_states(self)
env_class = _env_class
env = env_class(
env_params=env_params, sim_params=sim_params, scenario=scenario)
# create a Experiment object with the "rl_actions" method as
# described in the inputs. Note that the state may not be that which is
# specified by the environment.
exp = Experiment(env=env)
# run the experiment and return the reward
res = exp.run(
num_runs=NUM_RUNS,
num_steps=env.env_params.horizon,
rl_actions=_get_actions)
return np.mean(res["returns"]), np.std(res["returns"])
def bottleneck_example(flow_rate, horizon, restart_instance=False,
render=None):
"""
Perform a simulation of vehicles on a bottleneck.
Parameters
----------
flow_rate : float
total inflow rate of vehicles into the bottleneck
horizon : int
time horizon
restart_instance: bool, optional
whether to restart the instance upon reset
render: bool, optional
specifies whether to use the gui during execution
Returns
-------
exp: flow.core.experiment.Experiment
A non-rl experiment demonstrating the performance of human-driven
vehicles on a bottleneck.
"""
if render is None:
render = False
sim_params = AimsunParams(
sim_step=0.5,
render=render,
restart_instance=restart_instance)
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
num_vehicles=1)
additional_env_params = {
"target_velocity": 40,
"max_accel": 1,
"max_decel": 1,
"lane_change_duration": 5,
"add_rl_if_exit": False,
"disable_tb": DISABLE_TB,
"disable_ramp_metering": DISABLE_RAMP_METER
}
env_params = EnvParams(
horizon=horizon, additional_params=additional_env_params)
inflow = InFlows()
inflow.add(
veh_type="human",
edge="1",
vehsPerHour=flow_rate,
departLane="random",
departSpeed=10)
traffic_lights = TrafficLightParams()
if not DISABLE_TB:
traffic_lights.add(node_id="2")
if not DISABLE_RAMP_METER:
traffic_lights.add(node_id="3")
additional_net_params = {"scaling": SCALING, "speed_limit": 30/3.6}
net_params = NetParams(
inflows=inflow,
no_internal_links=False,
additional_params=additional_net_params)
initial_config = InitialConfig(
spacing="random",
min_gap=5,
lanes_distribution=float("inf"),
edges_distribution=["2", "3", "4", "5"])
scenario = BottleneckScenario(
name="bay_bridge_toll",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights)
env = BottleneckEnv(env_params, sim_params, scenario, simulator='aimsun')
return Experiment(env)
from flow.core.params import NetParams
net_params = NetParams(additional_params={
'length': 500,
'lanes': 3,
'speed_limit': 30,
'resolution': 40
})
from flow.core.params import InitialConfig
initial_config = InitialConfig(spacing="uniform", perturbation=1)
from flow.core.params import TrafficLightParams
traffic_lights = TrafficLightParams()
from flow.envs.ring.accel import AccelEnv
from flow.core.params import SumoParams
sim_params = SumoParams(sim_step=0.1, render=True, emission_path='data')
from flow.envs.ring.accel import ADDITIONAL_ENV_PARAMS
print(ADDITIONAL_ENV_PARAMS)
from flow.core.params import EnvParams
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
def variable_lanes_exp_setup(sim_params=None,
vehicles=None,
env_params=None,
net_params=None,
initial_config=None,
traffic_lights=None):
"""
Create an environment and network variable-lane ring road.
Each edge in this network can have a different number of lanes. Used for
test purposes.
Parameters
----------
sim_params : flow.core.params.SumoParams
sumo-related configuration parameters, defaults to a time step of 0.1s
and no sumo-imposed failsafe on human or rl vehicles
vehicles : Vehicles type
vehicles to be placed in the network, default is one vehicles with an
IDM acceleration controller and ContinuousRouter routing controller.
env_params : flow.core.params.EnvParams
environment-specific parameters, defaults to a environment with no
failsafes, where other parameters do not matter for non-rl runs
net_params : flow.core.params.NetParams
network-specific configuration parameters, defaults to a figure eight
with a 30 m radius
initial_config : flow.core.params.InitialConfig
specifies starting positions of vehicles, defaults to evenly
distributed vehicles across the length of the network
traffic_lights: flow.core.params.TrafficLightParams
traffic light signals, defaults to no traffic lights in the network
"""
logging.basicConfig(level=logging.WARNING)
if sim_params is None:
# set default sim_params configuration
sim_params = SumoParams(sim_step=0.1, render=False)
if vehicles is None:
# set default vehicles configuration
vehicles = VehicleParams()
vehicles.add(veh_id="idm",
acceleration_controller=(IDMController, {}),
car_following_params=SumoCarFollowingParams(
speed_mode="aggressive", ),
routing_controller=(ContinuousRouter, {}),
num_vehicles=1)
if env_params is None:
# set default env_params configuration
additional_env_params = {
"target_velocity": 8,
"max_accel": 1,
"max_decel": 1,
"sort_vehicles": False
}
env_params = EnvParams(additional_params=additional_env_params)
if net_params is None:
# set default net_params configuration
additional_net_params = {
"length": 230,
"lanes": 1,
"speed_limit": 30,
"resolution": 40
}
net_params = NetParams(additional_params=additional_net_params)
if initial_config is None:
# set default initial_config configuration
initial_config = InitialConfig()
if traffic_lights is None:
# set default to no traffic lights
traffic_lights = TrafficLightParams()
flow_params = dict(
# name of the experiment
exp_tag="VariableLaneRingRoadTest",
# name of the flow environment the experiment is running on
env_name=AccelEnv,
# name of the network class the experiment is running on
network=VariableLanesNetwork,
# simulator that is used by the experiment
simulator='traci',
# sumo-related parameters (see flow.core.params.SumoParams)
sim=sim_params,
# environment related parameters (see flow.core.params.EnvParams)
env=env_params,
# network-related parameters (see flow.core.params.NetParams and the
# network's documentation or ADDITIONAL_NET_PARAMS component)
net=net_params,
# vehicles to be placed in the network at the start of a rollout (see
# flow.core.params.VehicleParams)
veh=vehicles,
# parameters specifying the positioning of vehicles upon initialization/
# reset (see flow.core.params.InitialConfig)
initial=initial_config,
# traffic lights to be introduced to specific nodes (see
# flow.core.params.TrafficLightParams)
tls=traffic_lights,
)
# create the network
network = VariableLanesNetwork(name="VariableLaneRingRoadTest",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights)
# create the environment
env = AccelEnv(env_params=env_params,
sim_params=sim_params,
network=network)
# reset the environment
env.reset()
return env, network, flow_params
def make_create_env(params, version=0, render=None):
"""Create a parametrized flow environment compatible with OpenAI gym.
This environment creation method allows for the specification of several
key parameters when creating any flow environment, including the requested
environment and scenario classes, and the inputs needed to make these
classes generalizable to networks of varying sizes and shapes, and well as
varying forms of control (e.g. AVs, automated traffic lights, etc...).
This method can also be used to recreate the environment a policy was
trained on and assess it performance, or a modified form of the previous
environment may be used to profile the performance of the policy on other
types of networks.
Parameters
----------
params : dict
flow-related parameters, consisting of the following keys:
- exp_tag: name of the experiment
- env_name: name of the flow environment the experiment is running on
- scenario: name of the scenario class the experiment uses
- simulator: simulator that is used by the experiment (e.g. aimsun)
- sim: simulation-related parameters (see flow.core.params.SimParams)
- env: environment related parameters (see flow.core.params.EnvParams)
- net: network-related parameters (see flow.core.params.NetParams and
the scenario's documentation or ADDITIONAL_NET_PARAMS component)
- veh: vehicles to be placed in the network at the start of a rollout
(see flow.core.vehicles.Vehicles)
- initial (optional): parameters affecting the positioning of vehicles
upon initialization/reset (see flow.core.params.InitialConfig)
- tls (optional): traffic lights to be introduced to specific nodes
(see flow.core.params.TrafficLightParams)
version : int, optional
environment version number
render : bool, optional
specifies whether to use the gui during execution. This overrides
the render attribute in SumoParams
Returns
-------
function
method that calls OpenAI gym's register method and make method
str
name of the created gym environment
"""
exp_tag = params["exp_tag"]
env_name = params["env_name"] + '-v{}'.format(version)
module = __import__("flow.scenarios", fromlist=[params["scenario"]])
scenario_class = getattr(module, params["scenario"])
env_params = params['env']
net_params = params['net']
initial_config = params.get('initial', InitialConfig())
traffic_lights = params.get("tls", TrafficLightParams())
def create_env(*_):
sim_params = deepcopy(params['sim'])
vehicles = deepcopy(params['veh'])
scenario = scenario_class(
name=exp_tag,
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights,
)
# accept new render type if not set to None
sim_params.render = render or sim_params.render
# check if the environment is a single or multiagent environment, and
# get the right address accordingly
single_agent_envs = [env for env in dir(flow.envs)
if not env.startswith('__')]
if params['env_name'] in single_agent_envs:
env_loc = 'flow.envs'
else:
env_loc = 'flow.multiagent_envs'
try:
register(
id=env_name,
entry_point=env_loc + ':{}'.format(params["env_name"]),
kwargs={
"env_params": env_params,
"sim_params": sim_params,
"scenario": scenario,
"simulator": params['simulator']
})
except Exception:
pass
return gym.envs.make(env_name)
return create_env, env_name
return initial, net
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
routing_controller=(GridRouter, {}),
car_following_params=SumoCarFollowingParams(
min_gap=2.5,
decel=7.5, # avoid collisions at emergency stops
),
num_vehicles=tot_cars)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GrGr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yryr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rGrG"
def grid_example(render=None):
"""
Perform a simulation of vehicles on a grid.
Parameters
----------
render: bool, optional
specifies whether to use the gui during execution
Returns
-------
exp: flow.core.experiment.Experiment
A non-rl experiment demonstrating the performance of human-driven
vehicles and balanced traffic lights on a grid.
"""
v_enter = 10
inner_length = 300
long_length = 500
short_length = 300
n_rows = 2
n_columns = 3
num_cars_left = 20
num_cars_right = 20
num_cars_top = 20
num_cars_bot = 20
tot_cars = (num_cars_left + num_cars_right) * n_columns \
+ (num_cars_top + num_cars_bot) * n_rows
grid_array = {
"short_length": short_length,
"inner_length": inner_length,
"long_length": long_length,
"row_num": n_rows,
"col_num": n_columns,
"cars_left": num_cars_left,
"cars_right": num_cars_right,
"cars_top": num_cars_top,
"cars_bot": num_cars_bot
}
sim_params = SumoParams(sim_step=0.1, render=True)
if render is not None:
sim_params.render = render
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
routing_controller=(GridRouter, {}),
num_vehicles=tot_cars)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GrGrGrGrGrGr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yryryryryryr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rGrGrGrGrGrG"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "ryryryryryry"
}]
tl_logic.add("center0", phases=phases, programID=1)
tl_logic.add("center1", phases=phases, programID=1)
tl_logic.add("center2", phases=phases, programID=1, tls_type="actuated")
additional_net_params = {
"grid_array": grid_array,
"speed_limit": 35,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
if USE_INFLOWS:
initial_config, net_params = get_flow_params(
col_num=n_columns,
row_num=n_rows,
additional_net_params=additional_net_params)
else:
initial_config, net_params = get_non_flow_params(
enter_speed=v_enter,
add_net_params=additional_net_params)
scenario = SimpleGridScenario(
name="grid-intersection",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
env = AccelEnv(env_params, sim_params, scenario)
return Experiment(env)
flow_rate = 2000 * SCALING
# percentage of flow coming out of each lane
inflow = InFlows()
inflow.add(veh_type="rl",
edge="1",
vehs_per_hour=flow_rate * AV_FRAC,
departLane="random",
departSpeed=10)
inflow.add(veh_type="human",
edge="1",
vehs_per_hour=flow_rate * (1 - AV_FRAC),
departLane="random",
departSpeed=10)
traffic_lights = TrafficLightParams()
if not DISABLE_TB:
traffic_lights.add(node_id="2")
if not DISABLE_RAMP_METER:
traffic_lights.add(node_id="3")
additional_net_params = {"scaling": SCALING, "speed_limit": 23}
net_params = NetParams(inflows=inflow,
no_internal_links=False,
additional_params=additional_net_params)
flow_params = dict(
# name of the experiment
exp_tag="bottleneck_2",
# name of the flow environment the experiment is running on
# specify vehicle parameters to be added
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
acceleration_controller=(SimCarFollowingController, {}),
car_following_params=SumoCarFollowingParams(
min_gap=2.5,
decel=7.5, # avoid collisions at emergency stops
speed_mode="right_of_way",
),
routing_controller=(GridRouter, {}),
num_vehicles=0)
# Set up traffic light parameters
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GrGr"
}, {
"duration": "4",
"minDur": "3",
"maxDur": "6",
"state": "yryr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rGrG"
def variable_lanes_exp_setup(sim_params=None,
vehicles=None,
env_params=None,
net_params=None,
initial_config=None,
traffic_lights=None):
"""
Create an environment and scenario variable-lane ring road.
Each edge in this scenario can have a different number of lanes. Used for
test purposes.
Parameters
----------
sim_params : flow.core.params.SumoParams
sumo-related configuration parameters, defaults to a time step of 0.1s
and no sumo-imposed failsafe on human or rl vehicles
vehicles : Vehicles type
vehicles to be placed in the network, default is one vehicles with an
IDM acceleration controller and ContinuousRouter routing controller.
env_params : flow.core.params.EnvParams
environment-specific parameters, defaults to a environment with no
failsafes, where other parameters do not matter for non-rl runs
net_params : flow.core.params.NetParams
network-specific configuration parameters, defaults to a figure eight
with a 30 m radius and "no_internal_links" set to False
initial_config : flow.core.params.InitialConfig
specifies starting positions of vehicles, defaults to evenly
distributed vehicles across the length of the network
traffic_lights: flow.core.params.TrafficLightParams
traffic light signals, defaults to no traffic lights in the network
"""
logging.basicConfig(level=logging.WARNING)
if sim_params is None:
# set default sim_params configuration
sim_params = SumoParams(sim_step=0.1, render=False)
if vehicles is None:
# set default vehicles configuration
vehicles = VehicleParams()
vehicles.add(veh_id="idm",
acceleration_controller=(IDMController, {}),
car_following_params=SumoCarFollowingParams(
speed_mode="aggressive", ),
routing_controller=(ContinuousRouter, {}),
num_vehicles=1)
if env_params is None:
# set default env_params configuration
additional_env_params = {
"target_velocity": 8,
"max_accel": 1,
"max_decel": 1,
"sort_vehicles": False
}
env_params = EnvParams(additional_params=additional_env_params)
if net_params is None:
# set default net_params configuration
additional_net_params = {
"length": 230,
"lanes": 1,
"speed_limit": 30,
"resolution": 40
}
net_params = NetParams(additional_params=additional_net_params)
if initial_config is None:
# set default initial_config configuration
initial_config = InitialConfig()
if traffic_lights is None:
# set default to no traffic lights
traffic_lights = TrafficLightParams()
# create the scenario
scenario = VariableLanesScenario(name="VariableLaneRingRoadTest",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=traffic_lights)
# create the environment
env = AccelEnv(env_params=env_params,
sim_params=sim_params,
scenario=scenario)
# reset the environment
env.reset()
return env, scenario
data_file = 'flow/core/kernel/network/data.json'
with open(os.path.join(config.PROJECT_PATH, data_file)) as f:
data = json.load(f)
# export the data from the dictionary
veh_types = data['vehicle_types']
osm_path = data['osm_path']
if data['inflows'] is not None:
inflows = InFlows()
inflows.__dict__ = data['inflows'].copy()
else:
inflows = None
if data['traffic_lights'] is not None:
traffic_lights = TrafficLightParams()
traffic_lights.__dict__ = data['traffic_lights'].copy()
else:
traffic_lights = None
# generate the network
if osm_path is not None:
generate_net_osm(osm_path, inflows, veh_types)
edge_osm = {}
section_type = model.getType("GKSection")
for types in model.getCatalog().getUsedSubTypesFromType(section_type):
for s in types.itervalues():
s_id = s.getId()
num_lanes = s.getNbFullLanes()
length = s.length2D()
def grid1_baseline(num_runs, render=True):
"""Run script for the grid1 baseline.
Parameters
----------
num_runs : int
number of rollouts the performance of the environment is evaluated
over
render: bool, optional
specifies whether to the gui during execution
Returns
-------
flow.core.experiment.Experiment
class needed to run simulations
"""
exp_tag = flow_params['exp_tag']
sim_params = flow_params['sim']
vehicles = flow_params['veh']
env_params = flow_params['env']
net_params = flow_params['net']
initial_config = flow_params.get('initial', InitialConfig())
# define the traffic light logic
tl_logic = TrafficLightParams(baseline=False)
phases = [{
'duration': '31',
'minDur': '5',
'maxDur': '45',
"state": "GrGr"
}, {
'duration': '2',
'minDur': '2',
'maxDur': '2',
"state": "yryr"
}, {
'duration': '31',
'minDur': '5',
'maxDur': '45',
"state": "rGrG"
}, {
'duration': '2',
'minDur': '2',
'maxDur': '2',
"state": "ryry"
}]
for i in range(N_ROWS * N_COLUMNS):
tl_logic.add('center' + str(i),
tls_type='actuated',
phases=phases,
programID=1)
# modify the rendering to match what is requested
sim_params.render = render
# set the evaluation flag to True
env_params.evaluate = True
# import the network class
module = __import__('flow.networks', fromlist=[flow_params['network']])
network_class = getattr(module, flow_params['network'])
# create the network object
network = network_class(name=exp_tag,
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
# import the environment class
module = __import__('flow.envs', fromlist=[flow_params['env_name']])
env_class = getattr(module, flow_params['env_name'])
# create the environment object
env = env_class(env_params, sim_params, network)
exp = Experiment(env)
results = exp.run(num_runs, env_params.horizon)
total_delay = np.mean(results['returns'])
return total_delay
def para_produce_rl(HORIZON=3000):
# Create default environment parameters
env_params = EnvParams()
# Vehicle definition
vehicles = VehicleParams()
num_vehicles = 1
vehicles.add(
veh_id="human",
routing_controller=(GridRouter, {}),
lane_change_controller=(SimLaneChangeController, {}),
car_following_params=SumoCarFollowingParams(
min_gap=2.5,
decel=7.5, # avoid collisions at emergency stops
),
lane_change_params=SumoLaneChangeParams(
lane_change_mode=1621,
),
num_vehicles=num_vehicles)
# whether to allow turns at intersections
ALLOW_TURNS = False
# initialize traffic lights, used when you want define your own traffic lights
tl_logic = TrafficLightParams(baseline=False) # To see static traffic lights in action, the `TrafficLightParams` object should be instantiated with `baseline=False`
# when use off_ramp_grid.net.xml file, you should use a phase state example as "GGGgrrrrGGGgrrrr"
# when use off_ramp_grid_turn.net.xml file, you should use a phase state example as "GGGggrrrrrGGGggrrrrr"
if ALLOW_TURNS:
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
# for actuated traffic lights, you can add these optional values below
# "maxGap": int, describes the maximum time gap between successive vehicle sthat will cause the current phase to be prolonged
# "detectorGap": int, determines the time distance between the (automatically generated) detector and the stop line in seconds
# "showDetectors": bool, toggles whether or not detectors are shown in sumo-gui
"state": "GGGggrrrrrGGGggrrrrr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yyyyyrrrrryyyyyrrrrr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rrrrrGGGggrrrrrGGGgg"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "rrrrryyyyyrrrrryyyyy"
}]
tl_logic.add("center0", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center1", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center2", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center3", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
else:
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
# for actuated traffic lights, you can add these optional values below
# "maxGap": int, describes the maximum time gap between successive vehicle sthat will cause the current phase to be prolonged
# "detectorGap": int, determines the time distance between the (automatically generated) detector and the stop line in seconds
# "showDetectors": bool, toggles whether or not detectors are shown in sumo-gui
"state": "GGGgrrrrGGGgrrrr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yyyyrrrryyyyrrrr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rrrrGGGgrrrrGGGg"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "rrrryyyyrrrryyyy"
}]
# THIS IS A BUG THAT I DON'T KNOW WHY IT HAPPENS!!!!!!
phase0 = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GGrrGGrrGGrrGGrr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yyrryyrryyrryyrr"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rrGGrrGGrrGGrrGG"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "rryyrryyrryyrryy"
}]
tl_logic.add("center0", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center1", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center2", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
tl_logic.add("center3", phases=phases, programID=1, detectorGap=1,tls_type="actuated")
flow_params = dict(
exp_tag='offramp_multiagent_inflow_1.0_speed_20',
env_name=MultiTrafficLightGridPOEnv,
network=offRampGrid,
simulator='traci',
sim=SumoParams(
sim_step=0.1,
render=False,
#emission_path='./data',
restart_instance=True,
),
env=EnvParams(
horizon=3000, additional_params=ADDITIONAL_ENV_PARAMS.copy(),
),
net=net_params,
veh=vehicles,
initial=initial_config,
# used when you define your own traffic lights
#tls=tl_logic,
)
#flow_params['env'].horizon = HORIZON
return flow_params
def __init__(self,
name,
vehicles,
net_params,
initial_config=InitialConfig(),
traffic_lights=TrafficLightParams()):
"""Instantiate the base scenario class.
Attributes
----------
name : str
A tag associated with the scenario
vehicles : flow.core.params.VehicleParams
see flow/core/params.py
net_params : flow.core.params.NetParams
see flow/core/params.py
initial_config : flow.core.params.InitialConfig
see flow/core/params.py
traffic_lights : flow.core.params.TrafficLightParams
see flow/core/params.py
"""
# Invoke serializable if using rllab
if Serializable is not object:
Serializable.quick_init(self, locals())
self.orig_name = name # To avoid repeated concatenation upon reset
self.name = name + time.strftime('_%Y%m%d-%H%M%S') + str(time.time())
self.vehicles = vehicles
self.net_params = net_params
self.initial_config = initial_config
self.traffic_lights = traffic_lights
# specify routes vehicles can take
self.routes = self.specify_routes(net_params)
if net_params.template is None and net_params.osm_path is None:
# specify the attributes of the nodes
self.nodes = self.specify_nodes(net_params)
# collect the attributes of each edge
self.edges = self.specify_edges(net_params)
# specify the types attributes (default is None)
self.types = self.specify_types(net_params)
# specify the connection attributes (default is None)
self.connections = self.specify_connections(net_params)
# this is to be used if file paths other than the the network geometry
# file is specified
elif type(net_params.template) is dict:
if 'rou' in net_params.template:
veh, rou = self._vehicle_infos(net_params.template['rou'])
vtypes = self._vehicle_type(net_params.template.get('vtype'))
cf = self._get_cf_params(vtypes)
lc = self._get_lc_params(vtypes)
# add the vehicle types to the VehicleParams object
for t in vtypes:
vehicles.add(veh_id=t,
car_following_params=cf[t],
lane_change_params=lc[t],
num_vehicles=0)
# add the routes of the vehicles that will be departed later
# under the name of the vehicle. This will later be identified
# by k.vehicles._add_departed
self.routes = rou
# vehicles to be added with different departure times
self.template_vehicles = veh
self.types = None
self.nodes = None
self.edges = None
self.connections = None
# osm_path or template as type str
else:
self.nodes = None
self.edges = None
self.types = None
self.connections = None
# optional parameters, used to get positions from some global reference
self.edge_starts = self.specify_edge_starts()
self.internal_edge_starts = self.specify_internal_edge_starts()
self.intersection_edge_starts = [] # this will be deprecated
def da_yuan_example(render=None, use_inflows=True):
"""
Perform a simulation of vehicles on a traffic light grid.
Parameters
----------
render: bool, optional
specifies whether to use the gui during execution
use_inflows : bool, optional
set to True if you would like to run the experiment with inflows of
vehicles from the edges, and False otherwise
Returns
-------
exp: flow.core.experiment.Experiment
A non-rl experiment demonstrating the performance of human-driven
vehicles and balanced traffic lights on a traffic light grid.
"""
v_enter = 10
sim_params = SumoParams(sim_step=0.05, render=True)
if render is not None:
sim_params.render = render
vehicles = VehicleParams()
vehicles.add(
veh_id="human",
routing_controller=(GridRouter, {}),
car_following_params=SumoCarFollowingParams(
min_gap=2.5,
decel=7.5, # avoid collisions at emergency stops
),
num_vehicles=6)
env_params = EnvParams(additional_params=ADDITIONAL_ENV_PARAMS)
tl_logic = TrafficLightParams(baseline=False)
phases = [{
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "rrGGGG"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "rryyyy"
}, {
"duration": "31",
"minDur": "8",
"maxDur": "45",
"state": "GGrrrr"
}, {
"duration": "6",
"minDur": "3",
"maxDur": "6",
"state": "yyrrrr"
}]
tl_logic.add("inner0", phases=phases, programID=1)
tl_logic.add("inner1", phases=phases, programID=1)
tl_logic.add("inner2", phases=phases, programID=1, tls_type="actuated")
tl_logic.add("inner3", phases=phases, programID=1, tls_type="actuated")
additional_net_params = {
"speed_limit": 35,
"horizontal_lanes": 1,
"vertical_lanes": 1
}
if use_inflows:
initial_config, net_params = get_flow_params(
additional_net_params=additional_net_params)
else:
initial_config, net_params = get_non_flow_params(enter_speed=v_enter)
network = DaYuanNetwork(name="grid-intersection",
vehicles=vehicles,
net_params=net_params,
initial_config=initial_config,
traffic_lights=tl_logic)
env = AccelEnv(env_params, sim_params, network)
return Experiment(env)
