Traffic scenario

[FazelYU]

• [x] inject random traffic congestions
• [x] devise realistic demands
• [x] goal: large gap between [TTSP & TTSPWRR]

[FazelYU]

• [x] Uniform Demand

  • [x] Training phase demand: a uniform demanding with a maximum number of vehicles in the simulation
  • [x] Evaluation phase demand: the demand must be identical when testing for different baselines

• [x] Inject traffic congestion on the highway

[FazelYU]

• [x] Evaluation Phase Demand:

• get a number N as the sample size

• sample N trips in the network and write it to a demand file

• read from the demand file one by one in a cycle and insert it into the simulation as needed.

[FazelYU]

closes with issue #31

[FazelYU]

• [ ] Goal: a large gap between TTSP and TTSPWRR
• [ ] define congestion-prone paths:

  horizontal paths & highways each congestion prone path has a probability for congestion, and a congestion factor -[ ] during the "test phase" the "random seed" for congestion creation should be similar so that we can get comparable results