salvacorts / UGR-Metaheuristics

:seedling: Genetic Algorithm, Memetic Algorithms, GRASP, Simulated Annealing, Multi start search, Reiterated Local Search, Local Search, Greedy and randomized Greedy

A collection of different metaheuristics written in C++ to solve QAP problems.

Technical Paper comparing and explaining them (PDF in Spanish)

Available Algorithms

• Greedy
• Randomized Greedy
• Local Search
• Local Search with Don't loock Bits mask
• Genetic (Generational and Estacionary)
• Memetic (Based on Generational Genetic Algorithm)
• Simulated Annealing
• Basic Multistart Search
• GRASP
• Iterated Local Search
• Iterated Local Search using Simulated Annealing instead of Local search

Compilation

In order to compile with debug symbols:

cd build/
cmake -D CMAKE_BUILD_TYPE=Debug ..
make

To compile with optimization:

cd build/
cmake -D CMAKE_BUILD_TYPE=Release ..
make

Execution

./bin/main <problem> <random_seed>

Parameters:

• Problem: QAP problem definition, you can find some of them in instancias/
• Random Seed: Seed for random numbers generator, default is 7

Example:

./bin/main ./instancias/chr22a.dat 7