Effectiveness Evaluation of Memetics and Biogeography Algorithms Using Benchmark and Trans Computational Tasks of Combinatorial Optimization

摘要

The article discusses the elements of the theory of population metaheuristics. Original biogeographical and memetic algorithms for solving trans computational optimization problem are presented for the traveling salesman problem. Authors presented the method of biogeography and its modifications, as well as results of the comparative analysis of genetic, biogeographic and memetic algorithms. The authors conducted experimental verification of the effectiveness of the algorithms on known test functions. Experiments were carried out on certain benchmarks from the library TSPLIB. Efficiency, operating time and the diversity of the population were the criteria for comparison algorithms mentioned above.