ACOVRP

Web application to show the Ant System algorithm used to solve the Vehicle Routing Problem.

PARAMETERS DESCRIPTION 

There is no need to specify any parameters, they all have default values

• Iterations: Number of iterations of the algorithm. Default = 100
• Number of ants: The number of solutions constructed in each iterations. Default = 25
• Vehicle capacity: The maximum capacity of the vehicle (each node has a demand of around 10 units). Default depends on the problem
• q0: The probability of picking the best arc from a node. Default =0.9
• Beta: The importance of the distance of an arc in comparison with its pheromone level. Default = 2.3
• Pheromone change factor: The amount of pheromone an ant deposits in an arc. Default = 1
• Pheromone evaporation: The rate of evaporation of the pheromones. Default = 0.1
• Candidate size list: The n closest nodes to consider. Default = 50
• Multicolonies: Use a colony per vehicle route

Problems C1, C3, C4 are benchmark problems used in the paper. The parameters used are:

• 5000 iterations
• 25 ants
• The rest of the parameters are the default indicated above

Random problem creates 19 nodes (with demand =10) in random positions.