Solving CVRP with ACO
Minimizing Travel Cost for Complex Delivery Problems
This scenario involves the Capacitated Vehicle Routing Problem,
solved using the meta-heuristics algorithm Ant Colony Optimization. Basically, VRP is a network consisting of a number of nodes
(sometimes called cities) and arcs connecting one to all others along with the corresponding costs.
Mostly, the aim is to minimize the cost in visiting each customer once and only once. The term
"capacitated" is added due to some capacity constraints on the vehicles (vcap).
Enter the problem. Some company wants to deliver loads to a number of customers. In this case, we
have 24 nodes based on the location of Germany's train stations (don't ask why). The delivery
always starts from and ends at the depot, visiting a list of customers in other cities. And then
a number of questions arise:
- How do we minimize the travel cost in terms of distance?
- How many trucks are required?
- Which cities are visited by the truck #1, #2. etc.?
- depot: [0..23], def = 0
- vcap: [200..400], def = 400
There is a way to set all the demands, but I don't think you are ready for that. 😉
VCAP: 300 vol.
ACTIVE: 14 customers
- Kassel-Wilhelmshöhe (65 vol.)
- Düsseldorf Hbf (50 vol.)
- Aachen Hbf (25 vol.)
- Stuttgart Hbf (35 vol.)
- Hamburg Hbf (100 vol.)
- München Hbf (40 vol.)
- Leipzig Hbf (70 vol.)
- Nürnberg Hbf (60 vol.)
- Mannheim Hbf (55 vol.)
- Kiel Hbf (20 vol.)
- Mainz Hbf (70 vol.)
- Würzburg Hbf (65 vol.)
- Saarbrücken Hbf (70 vol.)
- Freiburg Hbf (50 vol.)
Tour 1
COST: 1715.296 km
LOAD: 300 vol.
- Würzburg Hbf | 65 vol.
- Stuttgart Hbf | 35 vol.
- Mannheim Hbf | 55 vol.
- Mainz Hbf | 70 vol.
- Aachen Hbf | 25 vol.
- Düsseldorf Hbf | 50 vol.
Tour 2
COST: 1978.053 km
LOAD: 290 vol.
- Leipzig Hbf | 70 vol.
- Nürnberg Hbf | 60 vol.
- München Hbf | 40 vol.
- Freiburg Hbf | 50 vol.
- Saarbrücken Hbf | 70 vol.
Tour 3
COST: 1165.889 km
LOAD: 185 vol.
- Kassel-Wilhelmshöhe | 65 vol.
- Hamburg Hbf | 100 vol.
- Kiel Hbf | 20 vol.
LOAD: 300 vol.
- Würzburg Hbf | 65 vol.
- Stuttgart Hbf | 35 vol.
- Mannheim Hbf | 55 vol.
- Mainz Hbf | 70 vol.
- Aachen Hbf | 25 vol.
- Düsseldorf Hbf | 50 vol.
LOAD: 290 vol.
- Leipzig Hbf | 70 vol.
- Nürnberg Hbf | 60 vol.
- München Hbf | 40 vol.
- Freiburg Hbf | 50 vol.
- Saarbrücken Hbf | 70 vol.
LOAD: 185 vol.
- Kassel-Wilhelmshöhe | 65 vol.
- Hamburg Hbf | 100 vol.
- Kiel Hbf | 20 vol.
#generations: 10 for global, 5 for local
#ants: 5 times #active_customers
ACO
Rel. importance of pheromones α = 1.0
Rel. importance of visibility β = 10.0
Trail persistance ρ = 0.5
Pheromone intensity Q = 10
See this wikipedia page to learn more.
NETWORK Depo: [1] Berlin Hbf | Number of cities: 24 | Total loads: 775 vol. | Vehicle capacity: 300 vol. Loads: [65, 0, 50, 0, 0, 25, 35, 0, 100, 40, 0, 70, 0, 60, 0, 0, 0, 55, 20, 70, 65, 70, 0, 50] ITERATION Generation: #1 Best cost: 5666.495 | Path: [1, 0, 2, 5, 19, 17, 6, 1, 11, 20, 13, 9, 23, 1, 8, 18, 21, 1] Best cost: 5000.313 | Path: [1, 2, 5, 19, 17, 6, 20, 1, 11, 13, 9, 21, 23, 1, 8, 18, 0, 1] Best cost: 4878.015 | Path: [1, 2, 5, 19, 17, 6, 20, 1, 11, 13, 9, 23, 21, 1, 18, 8, 0, 1] OPTIMIZING each tour... Current: [[1, 2, 5, 19, 17, 6, 20, 1], [1, 11, 13, 9, 23, 21, 1], [1, 18, 8, 0, 1]] [1] Cost: 1723.892 to 1715.296 | Optimized: [1, 20, 6, 17, 19, 5, 2, 1] [3] Cost: 1176.070 to 1165.889 | Optimized: [1, 0, 8, 18, 1] ACO RESULTS [1/300 vol./1715.296 km] Berlin Hbf -> Würzburg Hbf -> Stuttgart Hbf -> Mannheim Hbf -> Mainz Hbf -> Aachen Hbf -> Düsseldorf Hbf --> Berlin Hbf [2/290 vol./1978.053 km] Berlin Hbf -> Leipzig Hbf -> Nürnberg Hbf -> München Hbf -> Freiburg Hbf -> Saarbrücken Hbf --> Berlin Hbf [3/185 vol./1165.889 km] Berlin Hbf -> Kassel-Wilhelmshöhe -> Hamburg Hbf -> Kiel Hbf --> Berlin Hbf OPTIMIZATION RESULT: 3 tours | 4859.238 km.