Discord
@hash/rationally-navigating-london-uk
Simulation
6
0
Public

Person Agent in London UK

An example of a Person Agent navigating a city model of London.

An example of a Rational Agent navigating a city model of London.

This simulation combines two models - the first is the rational agent, which has preferences and goals it is trying to fulfill (ex. visit specific location). The second is a city model, in this case of London. The rational agent has preferences (ex. sleeping, eating), that it tries to fulfill by navigating London, and visiting specific destinations. When it visits a destination that can fulfill those preferences, it marks those as fulfilled and readjusts the preference weights to generate a new priority goal.

Preferences and the fulfillment criteria are set in the plan_generator.py action_table = { "hunger": [("move", ["destination"], [r_lon, r_lat])], "sleep": [("move", ["destination"], [h_lon, h_lat])], "work": [("move", ["destination"], [o_lon, o_lat])] }

The destinations for each preference are set in create_people.js for the first agent: "locs": { "home": { "lat": r.lng_lat[1], "lon": r.lng_lat[0] }, "office": { "lat": o.lng_lat[1], "lon": o.lng_lat[0] }, "restaurant": { "lon": -0.123081, "lat": 51.521960 }

See the rational agent model for more details on modifying the agent.