What is this?

This is a discrete event simulator written in javascript, made to be run under Node.js, simulating multiple separate queues. Each queue represents a city district which has it's own arrival rate (citizens calling the police), and service resources (number of police cars per district).

This repo also includes:

• Tests to find optimal distributions of resources (police cars) given finite resources (See "brute force programs"/get_best_for_19_cars.js and "brute force programs"/get_best_for_28_cars.js)
• A Presentation on the construction and usage of the simulator to find optimal results to problems in the case "COPS: Whatcha Gonna do?" Paper Here. We are essentialy trying to find optimal distribution(s) of finite resources over districts of varying service needs when using criterias such as average queue waiting time and server utilization.

Note

This was originally created for a business optimization / operations research course undertaken at university in 2017. The code has not been changed since the original implementation but if I get enough time at some point in the future I will push commits to address the short comings listed at the end of the presentation.

Installation:

$ npm install

Running the simulation:

To run the simulation for individual scenarios:

$ cd "brute force programs"
$ node simulation.js 5000 2 4 5 4 4

this will run the simulation (until 5000 discrete events are processed) with the distribution:

• District A: 2 cars,
• District B: 4 cars,
• ...
• District E: 4 cars

To get the exhaustive results i.e. test on all possible distributions when we have 19 cars to distribute over all districts

$ cd "brute force programs" 
$ node get_best_for_19_cars.js
or
$ node get_best_for_28_cars.js