| Describe 3 levels of planning horizons | Strategic: long-term 3-5 years+ and is reviewed annually
Tactical: mid-term 1-2 years and is reviewed quarterly
Operational: short-term, daily-1 year and is reviewed daily or weekly |
| Distribution planning considered.. | Strategic or tactical |
| Transport planning is considered.. | is considered a tactical or operational problem.
mostly outbound flow and rarely inbound flow (utgående transport ex)
We could plan inbound flow if it’s crucial for our production. |
| What issue does Traveling salesman problem (TSP) solve? | Solves the issue of minimizing the cost of having several customers that must be visited and then return home. |
| When is heuristics used instead of Traveling salesman problem? | Heuristics are used instead for whenever the problem gets too big |
| Name an issue with Traveling salesman problem | Considered “NP-hard” (computation time increases substantially as the number of customer increases) |
| what is Vehicle routing problem (VRP) and its aim? | Like TSP almost, there are a set of depots, a set of customers with known demands and vehicle/vehicles with capacities.
Also aims to minimize cost of delivery. |
| Most common formulations of VRP? | Can be formulated in many ways but the most common are Network models, Set Partitioning models, and Network flow models. |
| What differs between different formulations of VRP? | Each formulation has pros and cons, ex Understandability, Solution times. |
| Extensions of VRP could be? | Return of products
Pick-up and delivery (Pick something that’s required from one location or more and then to the customer)
Time windows, if there are specific times of delivery to the customer or the depot.
Arc routing, if you have to visit every arc/path at least once instead of every node/customer.
Split delivery
Multiple periods (produkten ska levereras nån gång i veckan exempelvis)
Dynamic VRP (kombinera efterfrågan med efterfrågan i realtid, ex vi ska leverera 8 i veckan men dem efterfrågar 10 helt plötsligt) |
| Common types of costs in different Variants of VRP formulations (vad för info kan vi söka efter) | Vehicle types
Fixed cost per tour (motsvarar administrativa kostnader)
Costs in proportion to transported distance (motsvarar löner, bränsle, slitage, etc)
Costs per customer visited (motsvarar administrativa kostnader)
Costs in proportion to delivered amount, loading and unloading cost (lastning- avlastningskostnader) |
| Alternative objective functions could be.. | Minimize driving distance
Minimise driving time
Minimise emissions
Minimise number of vehicles
Minimise delivery time to customer
Maximise number of visited customers
Maximize amount of delivered goods
Maximise fill rate of vehicles |
| What Simplifications of VRP could be made in real life.. | Experience (erfaren)
Fixed routes (Rutterna är redan kända)
Fixed areas of service per vehicles etc (kända platser där fordonen servas)
Optimization?
Simple heuristics, for example Clarke & Wright |
| Solutions principles for VRP.. | Optimization-based methods (lösningen ska vara optimeringsbaserad)
- Solve standard formulation/SPP with AMPL/CPLEX
- Relaxations & Branch-and-Bound
Often combined with heuristics for feasibility (Dvs kombineras med heuristiker för att det ska bli mer praktiskt)
- Limited on CPU or on optimality gap (Begränsas av våra datorers förmågor eller av optimality gap som är skillnaden/differensen mellan den bästa kända lösningen och det värdet som är vår lower bound. Lower bound värdet är typ det vi tror att vi kan uppnå, i detta fall är det minimeringsproblem) |
| Sweep heuristic characteristics.. | Easy and fast: can be run several times, with different start angles and directions.
Usually only for one-size fleet (samma kapacitet på varje fordon)
Using the underlying assumption of Euclidean distance (Euklidiskt avstånd är ett enkelt streck mellan två punkter, det antas alltså att det används) |
| Basic heuristic for VRP (solving heuristic).. | Assign a vehicle/rout for each single customer alone.
Compute the savings value of merging customers.
Make a list in decreasing savings value order.
Merge customers as long as possible |
