3049. Efficient decomposition methods for Large Scale, Multi-Period Log Truck Routing and Scheduling:Application to Canadian Forestry
Invited abstract in session MB-59: Routing decisions , stream Transportation.
Monday, 10:30-12:00Room: Liberty 1.14
Authors (first author is the speaker)
| 1. | Abdelhakim Abdellaoui
|
| Magi, Polytechnique Montreal | |
| 2. | Issmail El Hallaoui
|
| Math. and Ind. Eng., Polytechnique Montréal and GERAD | |
| 3. | Loubna Benabbou
|
| UQAR | |
| 4. | Francois Aubé
|
| Canmet Energy -Varennes |
Abstract
This work addresses the complex multi-period log-truck routing and scheduling problem encountered in the forest industry. We propose an improved mathematical formulation and decomposition approaches to efficiently solve large-scale instances of this problem. Given that timber harvesting operations in Canada extend far from processing facilities, efficient transportation is essential for economic viability and environmental sustainability.
Our research analyzes business rules within the forestry sector to develop a generalized framework for routing network design. We then formulate a comprehensive MILP, incorporating spatial and temporal constraints such as time windows, truck capacities, mill and harvest site accessibility, multi-period replenishment, and other resource constraints. To tackle the problem’s combinatorial complexity, we apply a preprocessing strategy to reduce the search space. We further develop primal decomposition methods, introducing a new Price-and-Branch approach and an adapted Relax-and-Fix strategy. Our computational experiments, conducted using historical data from a Canadian forest company, demonstrate the effectiveness of our approach, achieving near-optimal solutions (within a 2% gap of the lower bound) in less than 10 minutes for the weekly routing and scheduling problem. This research contributes to ongoing efforts to enhance operational efficiency, reduce environmental impact, and maintain the competitiveness of Canada’s forestry sector.
Keywords
- Combinatorial Optimization
- Column Generation
- Transportation
Status: accepted
Back to the list of papers