Parallel genetic algorithm based on GPU for solving quadratic assignment problem

One of the issues of combinatorial optimization is quadratic assignment problem (QAP). Solving this problem by using meta-heuristic algorithms to get good quality solution for average data takes a few minutes and for large data lasts for several hours. In this paper, to reduce the time to solve the problem of parallel genetic algorithm based on GPU (Graphics processing unit) is used. In addition, due to the problem of premature convergence of genetic algorithms, to improve results, some changes are applied on genetic algorithm. The results show that the proposed algorithm based on GPU gets more high-quality solutions in much less time than genetic algorithm based on CPU to solve the problem of QAP. In big problems, it acts 30X faster than base genetic algorithm.

Keywords:

Quadratic Assignment Problen(QAP) , Graphics Processing Unit(GPU) , CUDA , Genetics Algorithm(GA)