| Traffic Flow Theory And Simulation |
Article Index for Traffic |
Website Links For Traffic |
Information AboutTraffic Flow Theory And Simulation |
|
Just as general modeling and simulation found in other disciplines and engineering branches, transportation modeling and simulation is the mathematical abstraction of part of real world transportation systems and moving the mathematical model through time in a virtual environment. Many dimensions categorize transportation modeling and simulation, among which the level of modeling detail is widely known. Three levels of detail are available, i.e. macroscopic, mesoscopic, and microscopic, each of which progressively provides more modeling details. Currently under exploration is a fourth level of detail called nanoscopic transportation which, if implemented, could provide the finest modeling detail to transportation system operations. Representing one extreme of the spectrum, macroscopic simulation models roadway traffic as one-dimensional compressible fluid where disturbances in traffic flow propagate like waves. The fundamental law of macroscopic simulation states that, on a segment of the roadway, traffic flows in equals to traffic flows out plus any storage. Macroscopic simulation can achieve very large scale modeling with relatively low resolution. Examples of macroscopic simulators are KWaves, NETCELL, and KRONOS. Mesoscopic simulation is a tread-off between scale and resolution. A typical modeling technique adopted by mesoscopic simulation is Cellular Automata (CA) where time-space domain is partitioned into cells. Vehicles are modeled as particles which hop from one cell to another governed by some predetermined constraints. An example mesoscopic simulator is TRANSIMS. At the other end of the spectrum is microscopic simulation which provides high modeling resolution and also represents the state-of-the-art. Microscopic simulation models driver-vehicle units as particles, but the behavior of these particles is personalized by car-following, lane-changing, and gap-acceptance models. Examples of microscopic simulators are CORSIM, VISSIM, Paramics, AIMSUN, and HUTSIM. Another way to present these modeling philosophies is to make the following analogy. Suppose one is observing traffic 10,000 m above the ground, traffic behaves as a compressible fluid whose states (speed, flow, and density, etc.) propagate back and forth like waves. This is a scenario of macroscopic simulation. If one lowers to 3,000 m, the sense of waves recedes and a scene of particles emerges. A vehicle behaves as a particle hopping from one cell to another governed by some predetermined logics. This is a scenario of mesoscopic simulation. If one lowers even more to 1,000 m, the scene is dominated by moving particles which interact with each other so as to maintain safe positions in the traffic stream. This is a scenario of microscopic simulation as well as the state-of-the-art. Nanoscopic simulation is fundamentally different from the above modeling philosophies. Continuing with the above analogy, nanoscopic simulation provides a perspective as if one gets down to the ground and drives in the traffic stream. What one sees now is neither wave nor particle, but a complicated nanoscopic system consisting of drivers, vehicles, and environment (e.g. roadway, signs, signals, etc.). Drivers collect information and make control decisions in terms of steering, acceleration, and deceleration. Vehicles dynamically respond to drivers by executing the control decisions and moving on the ground. Feedback from vehicle dynamics, together with information from the environment, constitutes the basis for drivers to make control decisions in the next step and the above process goes on and on. Traffic operation is simply the movement and interaction of all vehicles in the system over time and space. This is a scenario of nanoscopic simulation. |
|
|