TRAFFIC EXTENDED (Original U. Wilensky, Extended by Forrest Sondahl)

WHAT IS IT? ----------- This project models the movement of cars on a (possibly multi-lane) highway. Each car follows a simple set of rules: it slows down if it sees a car close ahead, and speeds up if it doesn't see a car ahead. If it can change lanes to move around a slower car in its path, it does. The project demonstrates how traffic jams can form even without any accidents, broken bridges, or overturned trucks. No "centralized cause" is needed for a traffic jam to form. HOW TO USE IT ------------- Adjust the varous sliders to the appropriate settings you would like to test. Click on the SETUP button to set up the cars. Click on GO to start the cars moving. Note that they wrap the screen as they move, so the road is like a continuous loop. The SPEED-UP slider controls the rate at which cars accelerate when there are no cars ahead. When a car sees another car right in front, it matches that car's speed and then slows down a bit more. How much slower it goes than the car in front of it is controlled by the SLOW-DOWN slider. The SPEED-LIMIT slider controls the maximum speed that cars will go when they have open road ahead of them. The RANDOMNESS? switch determines whether cars will exhibit some randomness and variation in their speed. Randomness may cause cars to go marginally over the speed limit for short periods of time. The NUM-LANES slider controls the number of lanes. Changing its value while the simulation is running will not necessarily have the desired effect, so it is recommended that you change it and then run SETUP. THINGS TO NOTICE ---------------- Traffic jams can start from small "seeds." These cars start with random positions and random speeds. If some cars are clustered together, they will move slowly, causing cars behind them to slow down, and a traffic jam forms. Even though all of the cars are moving forward, the traffic jams tend to move backwards. This behavior is common in wave phenomena: the behavior of the group is often very different from the behavior of the individuals that make up the group. The plot shows three values as the model runs: - the fastest speed of any car (this doesn't exceed the speed limit!) - the slowest speed of any car - the speed of a single car (turtle 0), painted red so it can be watched. Notice not only the maximum and minimum, but also the variability -- the "jerkiness" of one vehicle. Sometimes it is possible to attain smoother flow (and greater average car speed) by cutting down the speed limit (probably to around 30), when randomness is turned on. THINGS TO TRY -------------- In this model there are three variables that can affect the tendency to create traffic jams: the initial NUMBER of cars, SPEED-UP, and SLOW-DOWN. Look for patterns in how the three settings affect the traffic flow. Which variable has the greatest effect? Do the patterns make sense? Do they seem to be consistent with your driving experiences? Set SLOW-DOWN to zero. What happens to the flow? Gradually increase SLOW-DOWN while the model runs. At what point does the flow "break down"? Does doubling the number of lanes double the capacity for traffic flow? Since each run is random, consider using the BehaviorSpace tool to run experiments that perform many runs, so that you can average the results for greater accuracy. EXTENDING THE MODEL ------------ Try other rules for speeding up and slowing down. Is the rule presented here realistic? Are there other rules that are more accurate or represent better driving strategies? In reality, different vehicles may follow different rules. Try giving different rules or speedup/slowdown values to some of the cars. Can one bad driver mess things up? What could you change to minimize the chances of traffic jams forming? What could you change to make traffic jams move forward rather than backward? Could the rules for lane-changing be more realistic? What would happen if drivers were required to keep right except to pass? Consider making a model with bidirectional traffic, where cars must pass by going into the other lane (when it is free from oncoming traffic). NETLOGO FEATURES ----------------- The plot shows both global values and the value for a single turtle, which helps one watch overall patterns and individual behavior at the same time. The WATCH command is used to make it easier to focus on the red car. RELATED MODELS --------------- "Traffic" (in StarLogoT) adds graphics, trucks, and a radar trap. "Gridlock" (a HubNet model which can be run as a participatory simulation) looks at traffic in a grid with many intersections. CREDITS AND REFERENCES ----------------------- This model was developed at the MIT Media Lab. See Resnick, M. (1994) "Turtles, Termites and Traffic Jams: Explorations in Massively Parallel Microworlds." Cambridge, Ma: MIT Press. Adapted to StarLogoT, 1997, as part of the Connected Mathematics Project. Adapted to NetLogo, 2000, as part of the Participatory Simulations Project. It was then extended to its present form by Forrest Sondahl for a class on Multi-Agent Modeling at Northwestern University in 2005. To refer to this model in academic publications, please use: Wilensky, U. (1997). NetLogo Traffic Basic model. http://ccl.northwestern.edu/netlogo/models/TrafficBasic. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL. In other publications, please use: Copyright 1997 by Uri Wilensky. All rights reserved. See http://ccl.northwestern.edu/netlogo/models/TrafficBasic for terms of use.