Traffic flow is defined as the number of vehicles that pass by a point in a given time period. This value is usually expressed as an hourly rate - for example, you can observe a flow of 500 vehicles per hour on a highway. where n is the number of vehicles that passed by a point during the time frame t.
I'm not certain I fully understood your query, so apologies if this reply isn't as focused as you might like.
The general factors to consider for modeling anything would be:
1. How variable is the outcome of interest in the population?
2. With what degree of precision would you like to estimate the form and function of a model?
3. How much risk are you willing to tolerate that your model (even with error bounds) will completely miss the mark?
4. Is the population finite or infinite?
5. How reliably can you measure the variable(s) in question?
and, of course,
6. What theory or rationale do you have to justify and evaluate a model? (Or, will this be an exercise akin to "data mining.")
Given answers to these aspects, I'm sure that there are likely many folks at your institution who could assist you in establishing a defensible sample size for your project (given your research aims).