Agent Based Bus Line

A python package for simulating a multi-lane bus route using agent-based modelling.

Installation

1. Download and unzip (or clone) this repository
2. Install python
3. Install the required python packages with pip install -r requirements.txt
4. Run the code with python ptips.py. Parameters you can adjust are contained at the bottom of that file.

Features

1. Closed loop traffic flow. All vehicles move in a single direction along the route. Cars have a probability of exiting the flow at green traffic lights and re-enter at red traffic lights to maintain a constant number of vehicles overall.
2. Multi-lane traffic. Busses remain in left-most lane at all times. Other vehicles change lanes as necessary.
3. Bus passenger routing system. Passengers arrive at bus stops over time. When loaded onto a bus, passengers are routed towards stops uniformly (i.e. all stops have equal weighting).

Parameters

All parameters are defined in a json5 file. See default.json for some default values:

Road system

• L = 1000 # circumference of circle (m)

Time marching

• t_max = 1e3 # maximum time (s)
• dt = 1e-1 # time increment (s)

Traffic properties

• initial_vehicle_spacing = 100 # (m/vehicle)
• speed_limit = 60 / 3.6 # maximum velocity (m/s)
• free_flowing_acceleration = 3 # typical vehicle acceleration (m/s^2)
• lanes = 2 # how many lanes

Bus system

• bus_fraction = 0.1 # what fraction of vehicles are busses (-)
• passenger_accumulation_rate = 0.1 # passengers arriving at a stop every second (passengers/s)
• passenger_ingress_egress_rate = 1 # how long to get on/off the bus (passengers/s)
• bus_max_capacity = 50 # maximum number of passengers on an individual bus (passengers/vehicle)
• bus_stop_traffic_light_offset = 0.5 # 0.1ish for just after the traffic lights, 0.9ish for just before traffic lights, 0.5 for in between (-)

Traffic light properties

• traffic_light_spacing = L / 4.0 # (m)
• traffic_light_period = 60 # (s)
• traffic_light_green_fraction = 0.5 # fraction of time it is green (-)
• car_entry_exit_probability = 0.1 # probability of moving to a different traffic light

Vehicle interaction properties

• stiffness = 1e4 # how much cars repel each other (also used for traffic lights, which are the same as stopped cars)
• sigma = 10 # typical stopping distance (m)

PTIPS stuff

• scheduled_velocity = 0.6 * speed_limit # how fast the busses are scheduled to move (m/s)
• ptips_delay_time = 10 # how much delay before PTIPS kicks in (s)
• ptips_capacity_threshold = 0.8 # how full should the busses be before ptips kicks in (-)