A Dynamic Messenger Problem Jan Fbry University of
A Dynamic Messenger Problem Jan Fábry University of Economics Prague fabry@vse. cz
Structure of presentation 1. Introduction 2. Static Messenger Problem 3. Messenger Problem with Time Windows 4. Dynamic Messenger Problem 5. Conclusion ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 2
1. Introduction Messenger problem - example 2 3 9 5 4 Depot 1 8 7 6 ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 3
2. Static Messenger Problem Requirements of ALL customers are known before a dispatcher starts to plan the route. ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 4
2. Static Messenger Problem Model assumptions i SOURCE (even nodes) i +1 DESTINATION (odd nodes) YES i i +1 NO ! ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 5
2. Static Messenger Problem Traveling Salesman Problem minimize subject to ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 6
2. Static Messenger Problem Messenger problem - example Optimal route 2 3 9 5 4 Depot 1 8 7 6 ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 7
3. Messenger Problem with TW minimize subject to {Time Window} ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 8
4. Dynamic Messenger Problem A new requirement for pick-up and delivery can occur during the travel. These two actions will be integrated into the current route. ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 9
4. Dynamic Messenger Problem Static problem Optimization Heuristic/metaheuristic A new requirement Dynamic problem Re-optimization Heuristic/metaheuristic ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 10
4. Dynamic Messenger Problem Insertion algorithm 11 10 Optimal route 2 3 9 4 Depot 1 8 5 7 6 ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 11
4. Dynamic Messenger Problem Insertion algorithm (1) (2) ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 12
4. Dynamic Messenger Problem Insertion algorithm Extension of the current route (1) (2) ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 13
4. Dynamic Messenger Problem Re-optimization minimize subject to {Set of non-visited locations} {Increasing sequence of locations of packages that have not been picked up} {Index of next location} ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 14
4. Dynamic Messenger Problem Re-optimization 11 10 Optimal route 2 3 9 4 Depot 1 8 5 7 6 ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 15
5. Conclusion • Heuristics & metaheuristics • Time windows • Multiple messengers - one depot - multiple depots • Capacity ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 16
Thank you for attention ☺ ____________________________________________ EURO XXII – European Conference on Operations Research, Prague 2007 17
- Slides: 17