Traffic Characteristics and Transportation Modeling
General Traffic Characteristics
The offer of a service cannot be booked for use during periods of high demand.
- The interaction of supply and demand takes over on an infrastructure, vehicles, and operating rules.
- The provision of infrastructure is expensive and of a discreet nature.
- The construction of infrastructure takes a long time (short-term inelastic supply).
- Operating a transportation system has externalities (pollution, accidents).
- Demand is highly qualitative and differential (goals, timetables, modes, by origin and destination).
- Demand is a derived demand and has a spatial component.
- Demand is dynamic in time (seasonal periodicity).
Transportation Modeling
A model is a representation of reality, an abstraction used for conceptual clarity, reducing variety and complexity to levels that can be adequately comprehended by analysis.
Types of Models
- Physical Models: Architectural models and wind tunnels are adequate but limited to the design aspect.
- Abstract Models: These are represented by symbols. See physical mechanisms. They are more useful for the planner.
Abstract Models (Mathematical Models)
These are related to algebraic models that contain two types of variables:
- Independent, Exogenous, Explanatory: Numerical values are determined outside the model.
- Endogenous or Dependent: The value results from the operation of the model.
The basic problem of modeling is estimating endogenous variables.
Predictive Models
Predictive models seek to find the right cause by positing variables, including type relations constant in time (to show greater productivity).
Regulatory Models
Regulatory models seek to optimize a given situation to improve the accessibility of an area (a problem is that they can be subjective).
Characteristics of Predictive Transport Models
- They consider the interplay between a large number of variables.
- They are associated with a wide range of objectives that operate simultaneously and, in some cases, even defy clear definition.
- They correspond to phenomena for which there are no good theories, and exchange systems are not yet well understood.
Development of a Model
- Purpose of Modeling: This largely determines the subsequent definitions.
- Time Available: The time available to produce the results also determines the level of detail.
- Modeling Area Boundaries: Is there a specific sector interaction with more remote areas?
- Resources: Financial resources, personnel, and time largely determine the scope of a model.
Choosing an Adequate Theory
It is possible that the same theories explain the same phenomenon. The model could fit the data with any theory. An additional problem is that different theories can lead to the same formal model whose parameters are interpreted differently.
Model Specification
- Structure of the model (simple structures selecting the alternatives independently).
- Functional form (linear or nonlinear functions).
- Specification of variables (which variables to use, in what form they will be used, and whether it will be possible to have that variable information).
- Zoning (which areas we want to represent, which are useful for the purpose of the study, which zones may be grouped).
- Representation of the network (passenger traffic lane use, how many and which connection points).
- Periodization (which period is of interest for modeling, such as days, weeks, or a year).
- Modes.
- Types of users (grouped by income or other interests).
Problems of Validation
- Replication accuracy of the base year data (used to calibrate the model).
- That the model has an adequate causal structure (e.g., motorization cup with banana sales, even if it serves no mathematical model).
- Provides no evidence at the time of the relations of the parameters (e.g., a hypothesis is correct 10 to 20 years from now).