/Simulated-Annealing-on-Travelling-Salesperson-Problem

Using Simulated Annealing on Travelling Salesperson Problem, tested against eil51 and kroA100 instances and achieving the best results of 437 and 26672 respectively

Primary LanguagePythonMIT LicenseMIT

Simulated Annealing on Travelling Salesperson Problem

Tested against eil51 and kroA100 instances with best results of 437 and 26672 respectively.

Initial temperature is dynamically set to when 94% of swappings are accepted.

Swapping Neighbors

Until the temperature reaches 20% of initial temperature, it uses the swapping technique of change a neighbor (each city has 2 diferrent neighbors) to the closest neighbor.

Below 20%, the whole city is swapped with another one randomly.