sybers / node-tsp-aco

TSP solver using Ant Colony Optimization

node-tsp-aco

TSP solver using Ant Colony Optimization

This project is a Node.js implementation of the Ant System algorithm proposed by M. Dorigo et al.

The main goal of this project was to produce a functional implementation of this algorithm and to test it against real world data to compare performance and accuracy of the algorithm.

There's also a Go implementation available here!

Installation

Clone this repository and move to the sources directory

$ git clone https://github.com/dzetah/node-tsp-aco.git node-tsp-aco
$ cd node-tsp-aco

Install dependencies

$ npm i

Transpile Typescript sources

$ npm run build

Run the TSP solver ! 🚀

$ node ./build/main.js

Usage

The following flags are available :

• -i, --interactive : Interactive mode.
• -a, --ants <value> : Number of ants (default 25).
• -A, --alpha <value> : Alpha value (default 1).
• -B, --beta <value> : Beta value (default 5).
• -g, --generations <value> : Number of generations (default 10).
• -e, --evaporation <value> Evaporation rate (default 0.5).
• -h, --help : displays the help message.

The Node.js version is not yet able to read .tsp files and uses the Oliver30 dataset by default.

Licence

MIT