Multi-population Genetic Algorithm for Rich Vehicle Routing Problems

Genetic Algorithm (GA) is the widely adopted meta-heuristic method for solving
Rich Vehicle Routing Problem (RVRP) due to its ability to find optimal solution even
for medium to large-scale problem in a reasonable time. However, genetic algorithm
is stochastic in nature and does not guarantee optimal solution in an application all
the time, a problem referred to as premature convergence in literature. In this pa�per we present Multi-population Genetic Algorithm for Rich Vehicle Routing Prob�lems (MPGA-RVRP) to provide diversity and delay premature convergence in GA by
making use of multiple populations that share potential solutions among each other
and evolve independently optimising only one objective. MPGA-RVRP is applied in
RVRP with three objectives:- total route distance, total route duration and total route
cost. Results from the experiments show that MPGA-RVRP performs better compared
to benchmark, Multi-objective Genetic Algorithm (MOGA). A web-based logistic sys�tem has also been developed as use case for MPGA-RVRP.