6533b86efe1ef96bd12cb556
RESEARCH PRODUCT
A multi-parametric evolution strategies algorithm for vehicle routing problems
Wout DullaertDavid MesterOlli Bräysysubject
Mathematical optimizationDynamic Source RoutingSDG 16 - PeaceComputer scienceEqual-cost multi-path routingEvolution strategiesArtificial IntelligenceVehicle routing problemVehicle routing problemHeuristicsDestination-Sequenced Distance Vector routingTriangular routingStatic routingDistribution managementPolicy-based routingSDG 16 - Peace Justice and Strong InstitutionsGeneral EngineeringPath vector protocol/dk/atira/pure/sustainabledevelopmentgoals/peace_justice_and_strong_institutionsJustice and Strong InstitutionsComputer Science ApplicationsDistance-vector routing protocolLink-state routing protocolMultipath routingHeuristicsAlgorithmdescription
Vehicle routing problems are at the heart of most decision support systems for real-life distribution problems. In vehicle routing problem a set of routes must be determined at lowest total cost for a number of resources (i.e. fleet of vehicles) located at one or several points (e.g. depots, warehouses) in order to efficiently service a number of demand or supply points. In this paper an efficient evolution strategies algorithm is developed for both capacitated vehicle routing problem and for vehicle routing problem with time window constraints. The algorithm is based on a new multi-parametric mutation procedure that is applied within the 1 + 1 evolution strategies algorithm. Computational testing on six real-life problems and 195 benchmark problems demonstrate that the suggested algorithm is efficient and highly competitive, improving or matching the current best-known solution in 42% of the test cases.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2007-02-01 | Expert systems with applications |