A Memetic Algorithm for Multi Objective Vehicle Routing Problem with Time Windows

Abstract

Vehicle routing problem is related to logistics and also called as VRP, which have many applications such as goods delivery, food transferring, supply chain management etc., VRP problem comes under the category of problems, which does not have an exact solution. So that, feasible solutions can be obtained first, later it is possible to find an optimal solution from them, which may further improved by using different algorithms. VRP is an abstraction of travelling sales person problem. VRPTW is a different category of VRP, which includes time constraints additionally, which looks like a multi-objective in the real world., a new memetic algorithm is defined for a central depot vehicle routing problem which achieves multiple objectives concurrently with specified time constraints (MA-MOVRPTW). Here concurrency represents that all objectives have same priority. The objective of MA-MOVRPTW is to find the least cost paths for vehicles to provide services to all customers along with the specified constraints. A local search algorithm existed to satisfy the above specified problem, but memetic algorithm is more efficient than the local search algorithm to produce better results. Finally, a memetic algorithm is provided for minimizing multiple objectives simultaneously with time windows.