On solving the vehicle routing problem using a flu-id genetic algorithm

Abstract

The article describes an approach to solving the vehicle routing problem based on a fluid genetic algorithm. Fluid genetic algorithms differ from classical ones by a more flexible coding scheme for solutions, which is relevant for problems with a complex solution structure. The article presents a mathematical formulation of the problem. The authors proposed two variants of encoding individuals, as well as an algorithm for recalculating the probabilities that form a chromosome in a fluid genetic algorithm. The proposed approach is compared with other existing approaches. The conducted research suggests that the use of fluid genetic algorithms to solve the vehicle routing problem is possible. The obtained results are correct, however, the algorithm works too slowly on a big data, and the resulting solution turns out to be significantly worse than the solution obtained by the classical genetic algorithm. The article discusses possible solutions to the problems encountered.