Here we model and simulate. Or simulate and model, depending on what we feel like.

To do list:

• make the cars start evenly spread across the road
• try out car distributions
• add logic to the movement of the cars
• implement physics from paper
• implement a measure of the traffic-jamness of the simulation, try to find relationships between variables and the amount of traffic jam generated
• mention limitations of our sim - cars don't move in parallel but one at a time
• push the zip file with older code to github @Andreea
• make a random slowdown parameter?
• make more lane change algorithms?
• make the road start with unbalanced densities
• mention that we are using no memory for the drivers like for lane changes etc
• talk about the acceleration to max speed ratio
• talk about random slowdown making graphs look chaotic, maybe more realistic
• try making random slowdown depend on something cause stochastic models and shit
• try making safety distance depend on speed - this sucks, try it with rand slowdowns

Meeting 1 notes

We discussed how to implement two-lane simulation using the existing code. We discussed potential ways to simulate lane switching, like:

• Switch based on relative lane speed
• Switch based on space in other lane if stopped (or set some speed cutoff)
• Switch based on lane density
  

We decided to save the code state as it is right now so that we have the simulation which was used to generate the 1D CA graphs.
We decided to not yet look at the other 2-lane CA implementation we found on github so that we can first try to implement our own without influence.
We decided to implement traffic jam metrics to be able to measure and graph the impact of changing constants on the outcome

• A measure of the fraction of time spent at 0 speed
• The fraction of time spent at max speed
• The fraction of maximum possible distance covered (hence taking into account not just stopping jams but also ones that slow the car down)