Journal of Heuristics, Journal Year: 2025, Volume and Issue: 31(2)
Published: April 22, 2025
Language: Английский
Citations
1Inverse Optimization for Routing Problems
Transportation Science, Journal Year: 2024, Volume and Issue: unknown
Published: July 17, 2024
We propose a method for learning decision makers’ behavior in routing problems using inverse optimization (IO). The IO framework falls into the supervised category and builds on premise that target is an optimizer of unknown cost function. This function to be learned through historical data, context problems, can interpreted as preferences makers. In this view, main contributions study are methodology with hypothesis function, loss stochastic first-order algorithm tailored problems. further test our approach Amazon Last Mile Routing Research Challenge, where goal learn models replicate human drivers, thousands real-world examples. Our final IO-learned model achieves score ranks second compared 48 qualified round challenge. examples results showcase flexibility potential proposed from decision-makers’ decisions History: paper has been accepted Transportation Science Special Issue TSL Conference 2023. Funding: work was supported by European Council [TRUST-949796].
Language: Английский
Citations
7Last mile delivery routing problem with some-day option
European Journal of Operational Research, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 1, 2025
Language: Английский
Citations
0Stork: A Flexible Optimization Tool for Solving Multiple Vehicle Routing Problem Variants
Transportation research procedia, Journal Year: 2025, Volume and Issue: 86, P. 258 - 265
Published: Jan. 1, 2025
Language: Английский
Citations
0An Iterative Sample Scenario Approach for the Dynamic Dispatch Waves Problem
Transportation Science, Journal Year: 2024, Volume and Issue: 58(4), P. 726 - 740
Published: March 19, 2024
A challenge in same-day delivery operations is that requests are typically not known beforehand, but instead revealed dynamically during the day. This uncertainty introduces a trade-off between dispatching vehicles to serve as soon they ensure timely and delaying decision consolidate routing decisions with future, currently unknown requests. In this paper, we study dynamic dispatch waves problem, problem which dispatched at fixed moments. At each moment, system operator must decide of how route these The operator’s goal minimize total cost while ensuring all served on time. We propose iterative conditional (ICD), an solution construction procedure based sample scenario approach. ICD iteratively solves scenarios classify be dispatched, postponed, or undecided. set undecided shrinks iteration until final made last iteration. develop two variants ICD: one variant thresholds, another similarity. significant strength it conceptually simple easy implement. simplicity does harm performance: through rigorous numerical experiments, show both efficiently navigate large state action spaces quickly converge high-quality solution. Finally, demonstrate threshold-based achieves excellent results instances from EURO Meets NeurIPS 2022 Vehicle Routing Competition, nearly matching performance winning machine learning–based strategy. History: paper has been accepted for Transportation Science Special Issue DIMACS Implementation Challenge: Problems. Funding: work was supported by TKI Dinalog, Topsector Logistics, Dutch Ministry Economic Affairs Climate Policy. Supplemental Material: online appendix available https://doi.org/10.1287/trsc.2023.0111 .
Language: Английский
Citations
3Last Mile Delivery Routing Problem with Some-Day Option
SSRN Electronic Journal, Journal Year: 2024, Volume and Issue: unknown
Published: Jan. 1, 2024
E-commerce retailers are challenged to maintain cost-efficiency and customer satisfaction while pursuing sustainability, especially in the last mile. In response, offering a range of delivery speeds, including same-day instant options. Faster deliveries, trending, often increase costs emissions due limited planning time reduced consolidation opportunities contrast, this paper proposes inclusion slower option, termed some-day. Slowing down process allows for greater shipment consolidation, achieving cost savings environmental goals simultaneously.We introduce dynamic stochastic some-day problem, which accounts latest day, windows, capacity limitations within multi-period framework. Our solution approach is based on addressing auxiliary prize-collecting vehicle routing problems with windows (PCVRPTW) daily basis, where prize reflects benefit promptly serving customer. We develop hybrid adaptive large neighborhood search granular insertion operators, outperforming existing metaheuristics PCVRPTWs. numerical study shows significant only small increases times compared an earliest policy.
Language: Английский
Citations
1Learn to Solve Vehicle Routing Problems Asap:A Neural Optimization Approach for Time-Constrained Vehicle Routing Problems with Finite Vehicle Fleet
Published: Jan. 1, 2024
Language: Английский
Citations
1An iterative sample scenario approach for the dynamic dispatch waves problem
arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown
Published: Jan. 1, 2023
A challenge in same-day delivery operations is that requests are typically not known beforehand, but instead revealed dynamically during the day. This uncertainty introduces a trade-off between dispatching vehicles to serve as soon they ensure timely delivery, and delaying decision consolidate routing decisions with future, currently unknown requests. In this paper, we study dynamic dispatch waves problem, problem which dispatched at fixed moments. At each moment, system operator must decide of dispatch, how route these The operator's goal minimize total cost while ensuring all served on time. We propose iterative conditional (ICD), an solution construction procedure based sample scenario approach. ICD iteratively solves scenarios classify be dispatched, postponed, or undecided. set undecided shrinks iteration until final made last iteration. develop two variants ICD: one variant thresholds, another similarity. significant strength it conceptually simple easy implement. simplicity does harm performance: through rigorous numerical experiments, show both efficiently navigate large state action spaces quickly converge high-quality solution. Finally, demonstrate threshold-based achieves excellent results instances from EURO meets NeurIPS 2022 vehicle competition, nearly matching performance winning machine learning-based strategy.
Language: Английский
Citations
1Coordinating Delivery Service Providers for Greener Two-Echelon Vehicle Routing Via Bilevel Optimization: Exact and Data-Driven Methods
Published: Jan. 1, 2024
Language: Английский
Citations
0The bulk of it: Simultaneous pickup and delivery of furniture
Published: Jan. 1, 2024
Language: Английский
Citations
0