This is my problem: I am planning camping trips for multiple people. Each person will be given a route. The routes may intersect at some points. They all start their trip simultaneously. Whenever two routes intersect, the person arriving the intersection point first must wait for the other person to arrive, after which both are allowed to continue their respective trips. The goal is to minimize the time it takes for the last person to finish his/her trip.
If the weather is to bad the travelers have to stop, but each traveler might have a different weather threshold before having to stop. I have historic weather data available for the last x years.
What is currently done is that a complete plan, namely a route for every traveler, is simulated by playing out what would happen each of the x years, then an average completion time over the x years is calculated. Hence, during the simulation no random numbers are generated as weather changes are given by the weather data.
- Is this a sensible way of doing it, given that it is very difficult to generate random but at the same time realistic weather scenarios?
- Would this be considered a discrete event simulation based approach?