A customized genetic algorithm for bi-objective routing in a dynamic network

The article presents a proposed customized genetic algorithm (CGA) to find the Pareto frontier for bi-objective integer linear programming (ILP) model of routing in dynamic network, where number nodes and edge weights vary over time. Utilizing hybrid method, CGA combines with (DP); it is fast alternative an ILP solver finding efficient solutions, particularly large dimensions. A non-dominated sorting (NSGA-II) used as base multi-objective evolutionary algorithm. Real data are target trajectories, from case study application surveillance boat measure greenhouse-gas emissions ships on Baltic sea. CGA’s performance evaluated comparison solutions terms accuracy computation efficiency. Results multiple runs indicate convergence frontier, considerable speed-up relative solver. stays hybridizing optimization DP methods together solving complex real-world problems.